About the Montreal Classification of Inflammatory Bowel Disease
Nothing here is a severity score — Montreal is a notation for recording a phenotype so that two clinicians describing the same patient produce the same string. Crohn's disease is coded on three axes: age at diagnosis (A1 ≤ 16 years, A2 17–40, A3 > 40), location (L1 ileal, L2 colonic, L3 ileocolonic, L4 isolated upper gastrointestinal), and behaviour (B1 non-stricturing non-penetrating, B2 stricturing, B3 penetrating), with a lower-case p appended for perianal disease. Ulcerative colitis uses two: extent (E1 proctitis, E2 left-sided, E3 extensive) and severity (S0 remission through S3 severe). The descriptors earn their place because several of them change what you do — B2 and B3 push toward earlier biologics and surgical input, and E2 or E3 sets the clock running on colorectal cancer surveillance.
Formula
Crohn's: A{1–3} L{1–4} B{1–3}[p] Ulcerative colitis: E{1–3} S{0–3}- A1 / A2 / A3
- Age at diagnosis: ≤ 16 years, 17–40 years, > 40 years. Fixed permanently at diagnosis and never revised.
- L1 / L2 / L3 / L4
- Ileal, colonic, ileocolonic, isolated upper gastrointestinal. L4 doubles as a modifier — L3+L4 is a valid description and was not expressible under Vienna.
- B1 / B2 / B3
- Non-stricturing non-penetrating, stricturing, penetrating. Hierarchical — penetrating overrides stricturing — and expected to progress with disease duration.
- p
- Perianal disease modifier, appended to the behaviour letter. Independent of B: B1p is a coherent and common phenotype.
- E1 / E2 / E3
- Proctitis, left-sided (to the splenic flexure), extensive (proximal to the splenic flexure). Recorded as the maximal extent ever documented.
- S0 / S1 / S2 / S3
- Clinical remission, mild, moderate, severe. Descriptive severity bands rather than a validated index.
- No arithmetic. Montreal produces a label, not a number, and there is no total to compute or threshold to cross.
- Behaviour is the only Crohn's axis that changes. Age at diagnosis is fixed for life, and location is largely stable — a documented change in location should prompt a review of whether the original assessment was complete rather than being recorded as evolution.
- Ulcerative colitis extent is the maximum ever documented, so a patient with previous pancolitis in endoscopic remission remains E3. Downgrading extent because the current scope was normal is a common and consequential error.
- Severity (S) is the one axis describing the present, and it is the weakest part of the classification — validated activity indices exist and should be used alongside it.
- Paediatric practice uses the Paris modification, which subdivides A1, splits E4 for pancolitis, adds growth failure, and allows both B2 and B3 to be recorded together.
Interpreting the result
For Crohn's disease, read behaviour first. B1 describes inflammatory disease without structural complication and supports a conventional step-up approach; B2 and B3 indicate that structural damage has occurred and shift the conversation toward earlier biologic therapy and early surgical involvement, since a fibrotic stricture will not respond to immunosuppression and a penetrating complication needs a plan rather than an escalation. The p modifier carries independent weight: perianal fistulising disease is a distinct therapeutic problem with its own evidence base, and it should not be inferred from or folded into B3. Location matters mainly for drug delivery and for what surveillance is relevant — extensive colonic involvement in Crohn's disease carries a colorectal cancer risk comparable to ulcerative colitis of similar extent. For ulcerative colitis, extent drives two decisions. E1 proctitis is usually a topical-therapy problem and carries no increased colorectal cancer risk; E2 and E3 need oral or systemic therapy, and both start colonoscopic surveillance approximately eight years from symptom onset. Severity should be read as a rough current descriptor and paired with a validated index before treatment decisions rest on it.
| Score | Band | What it means | Action |
|---|---|---|---|
| B1 (± p) | Non-stricturing, non-penetrating Crohn's | Inflammatory disease without structural complication. The phenotype most likely to respond to medical therapy alone | Conventional escalation guided by an activity index and objective inflammation; monitor for transition to B2 or B3 |
| B2 (± p) | Stricturing Crohn's | Fixed luminal narrowing. Fibrotic components will not respond to immunosuppression | Distinguish inflammatory from fibrotic stricture on imaging; involve surgery early and consider earlier biologic therapy |
| B3 (± p) | Penetrating Crohn's | Fistulising or perforating disease. Associated with a more aggressive course and higher surgical requirement | Cross-sectional imaging to define the tracks, exclude abscess before immunosuppression, and involve surgery early |
| E1 | Ulcerative proctitis | Inflammation limited to the rectum. No increased colorectal cancer risk attributable to extent | Topical therapy is first-line and often sufficient; surveillance colonoscopy is not indicated on the basis of extent |
| E2 | Left-sided ulcerative colitis | Extends to but not beyond the splenic flexure. Carries an increased colorectal cancer risk | Oral therapy alongside topical; begin colonoscopic surveillance approximately 8 years from symptom onset |
| E3 | Extensive ulcerative colitis / pancolitis | Extends proximal to the splenic flexure. Highest colorectal cancer risk and highest colectomy rate | Systemic therapy; surveillance from approximately 8 years after symptom onset, at intervals set by risk stratification |
Scroll the table sideways for every column.
What the Montreal IBD needs (7 inputs)
- Disease — Crohn's disease or ulcerative colitis
- The two diseases use entirely different axes, so this selection determines which descriptors apply. Montreal also acknowledges IBD unclassified, for colitis that cannot be assigned to either — a category the classification names but does not code.
- A — age at diagnosis (Crohn's)
- A1 is 16 years or under, A2 is 17 to 40, A3 is over 40. This is one of the changes Montreal made to the Vienna classification, which had used 40 as a single dividing line and had no paediatric category at all.
- L — disease location (Crohn's)
- L1 ileal, L2 colonic, L3 ileocolonic, L4 isolated upper gastrointestinal. L4 can also be added as a modifier to L1, L2 or L3 when upper gastrointestinal disease coexists with distal disease — under Vienna it was mutually exclusive, which forced a choice that misdescribed the patient.
- B — disease behaviour (Crohn's)
- B1 non-stricturing and non-penetrating, B2 stricturing, B3 penetrating. Behaviour is hierarchical: a patient with both a stricture and a fistula is B3. It is also the axis that changes over time, and the only one that should be expected to.
- p — perianal disease modifier (Crohn's)
- Appended to the behaviour category as a lower-case p, giving strings such as B1p or B3p. Montreal deliberately separated perianal disease from penetrating disease, because a patient can have perianal fistulising disease with no internal penetrating complication, and Vienna's coding conflated the two.
- E — disease extent (ulcerative colitis)
- E1 ulcerative proctitis, E2 left-sided or distal disease (up to the splenic flexure), E3 extensive disease or pancolitis (proximal to the splenic flexure). Extent is defined by the maximal proximal extent of macroscopic inflammation ever documented, not by the current endoscopic picture.
- S — severity (ulcerative colitis)
- S0 clinical remission, S1 mild, S2 moderate, S3 severe. The one Montreal axis that does describe current state, which is why it is the part of the classification most often superseded in practice by a formal activity index.
What it returns
- Montreal classification string
- For Crohn's disease, a three-part code with an optional p — for example A2L3B2p. For ulcerative colitis, a two-part code — for example E3S2.
- What the phenotype implies
- The management consequences that follow from the descriptors, particularly behaviour in Crohn's disease and extent in ulcerative colitis.
How it is calculated
Montreal is a consensus classification produced by a working party at the 2005 World Congress of Gastroenterology, revising the 1998 Vienna classification of Crohn's disease and extending the exercise to ulcerative colitis, which Vienna had not addressed. The method was expert consensus informed by the accumulated phenotyping literature rather than statistical derivation, which is appropriate for a vocabulary: the goal was agreement on definitions, not discovery of thresholds. The working party's stated ambition went further than the three axes that were adopted — the report is titled toward an integrated clinical, molecular and serological classification, and the intention was that genetic and serological markers would eventually be incorporated. That has not happened in routine practice, so what survives is the clinical layer, which is why the classification looks purely phenotypic despite its origins.
Facts & figures
| Element | Vienna (1998) | Montreal (2005) |
|---|---|---|
| Age at diagnosis | A1 < 40, A2 ≥ 40 | A1 ≤ 16, A2 17–40, A3 > 40 — a paediatric category added |
| Upper GI disease | L4 mutually exclusive with L1–L3 | L4 usable as a modifier alongside L1–L3 |
| Perianal disease | Folded into penetrating behaviour | Separate lower-case p modifier, independent of B |
| Ulcerative colitis | Not addressed | E1–E3 extent and S0–S3 severity |
Scroll the table sideways for every column.
Each change fixed a case the earlier scheme described wrongly rather than merely imprecisely — a 12-year-old and a 39-year-old shared a category, ileocolonic disease with gastric involvement could not be coded, and perianal fistulising disease without internal penetration had nowhere to go.
| Grade | Description |
|---|---|
| S0 — clinical remission | Asymptomatic |
| S1 — mild | Up to four stools daily (with or without blood), no systemic illness, normal inflammatory markers |
| S2 — moderate | More than four stools daily with minimal signs of systemic toxicity |
| S3 — severe | At least six bloody stools daily, pulse ≥ 90, temperature ≥ 37.5 °C, haemoglobin < 10.5 g/dL, ESR ≥ 30 mm/h |
Scroll the table sideways for every column.
The S3 criteria are recognisably Truelove and Witts. That is not a coincidence — Montreal adopted the existing severe-colitis definition rather than creating a new one, which is why acute severe ulcerative colitis is still assessed with Truelove and Witts at the bedside.
Evidence
Derivation — Working Party, Montreal World Congress 2005
2005Consensus classification produced by a working party of the 2005 Montreal World Congress of Gastroenterology, revising the 1998 Vienna classification of Crohn's disease and extending phenotyping to ulcerative colitis. Developed by expert consensus informed by the phenotyping literature rather than fitted to a cohort, which is the appropriate method for a descriptive vocabulary.
No sensitivity or discrimination statistics exist or apply — the classification defines terms rather than predicting outcomes. Its value rests on reproducibility of description and on the prognostic associations of the individual descriptors, particularly behaviour and extent.
Consensus commentary — Satsangi 2006
2006Published discussion of the Montreal classification's controversies, points of consensus and practical implications, in Gut.
Set out the rationale for the changes from Vienna, including the paediatric age category, the L4 modifier and the separation of perianal disease, and identified the ulcerative colitis severity axis as the least secure element.
Paediatric modification — Paris classification, 2011
2011Multinational paediatric consensus modifying Montreal for children, driven by evidence that paediatric-onset disease differs in location, behaviour and progression.
Subdivided A1 into A1a (0 to < 10 years) and A1b (10 to < 17 years), added E4 for pancolitis, permitted B2 and B3 to be recorded simultaneously, and added a growth-failure descriptor absent from Montreal.
How it compares
Montreal IBD vs Vienna classification
Montreal supersedes Vienna and should be used instead — the changes fixed phenotypes Vienna described incorrectly rather than merely coarsely.
Vienna coded Crohn's disease on the same three axes but drew them differently. It split age at 40 with no paediatric category, so a 12-year-old and a 39-year-old shared a group; it made L4 mutually exclusive with L1–L3, forcing a false choice for a patient with both gastric and ileocolonic disease; it folded perianal disease into penetrating behaviour, leaving perianal fistulising disease without internal penetration unrepresentable; and it did not address ulcerative colitis at all. Montreal corrected each. Vienna codes still appear in older records and in long-running cohort studies, so recognising them matters, but new classification should be Montreal.
Montreal IBD vs Paris classification
Paris is the paediatric version and should be used for childhood-onset disease — Montreal's single A1 category is too coarse for a population in which age of onset predicts phenotype.
The Paris modification subdivides A1 into A1a (0 to under 10 years) and A1b (10 to under 17), adds E4 for pancolitis, allows stricturing and penetrating behaviour to be recorded together rather than hierarchically, and adds a growth-failure descriptor that Montreal has no way to express. Each change reflects evidence that paediatric-onset inflammatory bowel disease differs in location, in rate of progression and in its consequences for development. Using Montreal in a paediatric clinic is not wrong so much as lossy — the descriptors that paediatric management turns on are the ones it cannot record.
Montreal IBD vs Mayo score / SCCAI / UCEIS
Not alternatives — Montreal records what the disease is and where it is, while the activity indices record how it is behaving today, and ulcerative colitis management needs both.
Montreal's E axis and the activity indices answer different questions and are frequently confused because both concern ulcerative colitis. Extent is a stable structural fact that determines the route of therapy and the surveillance obligation; activity is a fluctuating measure that determines whether current therapy is working. The Montreal S axis blurs this by describing severity, which is why it is the weakest part of the classification and why it has largely been displaced in practice — the Mayo score and SCCAI quantify symptomatic activity, UCEIS quantifies endoscopic activity, and all three are validated for measuring change in a way that S0 to S3 is not.
Montreal IBD vs Harvey-Bradshaw Index / CDAI
Complementary — Montreal fixes the Crohn's phenotype in the record, the activity indices track the disease week to week, and treatment decisions draw on both.
A patient described as A2L3B2p has a phenotype that predicts a more complicated course and argues for earlier biologic therapy and surgical involvement. Whether they need a change of treatment this month is a separate question, answered by the Harvey-Bradshaw Index or CDAI alongside objective markers of inflammation. The distinction has practical bite in stricturing disease, where a high activity index may reflect obstructive symptoms from fibrosis rather than active inflammation — the phenotype tells you to check which, and the activity index alone will mislead if you do not.
Pearls & pitfalls
- Montreal is not a severity score. There is no total, no threshold and no cut-off — a request for 'the Montreal score' is a request for a label.
- Ulcerative colitis extent is the maximum ever documented. A patient with previous pancolitis whose current scope is normal remains E3, and downgrading them removes their surveillance indication.
- Behaviour is hierarchical: a patient with both a stricture and a fistula is B3, not B2 and B3. Only the Paris paediatric modification permits recording both.
- The p modifier is independent of behaviour. B1p is common and coherent — perianal fistulising disease does not make a patient B3.
- L4 is both a category and a modifier. Isolated upper gastrointestinal disease is L4; upper gastrointestinal disease alongside ileocolonic disease is L3+L4, which Vienna could not express.
- Age at diagnosis is fixed permanently. It describes the disease at onset, not the patient now, and a 50-year-old diagnosed at 14 is A1 for life.
- Do not use the S axis as an activity measure for treatment decisions. Use the Mayo score, SCCAI, UCEIS or Truelove and Witts, which are validated for that purpose.
- Extensive colonic Crohn's disease carries a colorectal cancer risk broadly comparable to ulcerative colitis of similar extent — surveillance follows the extent of colonic involvement, not the diagnostic label.
- In children, use the Paris modification. Applying Montreal to a 7-year-old loses the age subdivision and the growth-failure descriptor that paediatric practice depends on.
Critical actions
- Record the full string in the problem list and the clinic letter, not buried in prose — the point of a notation is that it can be found later.
- Date the behaviour descriptor. B1 in 2019 and B3 in 2026 is a prognostically important trajectory that an undated label erases.
- In ulcerative colitis, record extent as the maximum ever documented and note where that assessment came from, so a later normal colonoscopy does not appear to contradict it.
- Start colonoscopic surveillance approximately eight years from symptom onset in E2 and E3 disease, and in Crohn's disease with comparable colonic extent.
- On finding a new stricture, image before escalating medical therapy — distinguishing inflammatory from fibrotic narrowing determines whether immunosuppression can work at all.
- Exclude an abscess before starting immunosuppression or a biologic in B3 disease.
- Involve colorectal surgery early in B2 and B3 disease rather than at the point of failure of medical therapy.
- Pair the classification with a validated activity index at every visit where a treatment decision is being made.
- Use the Paris modification for patients diagnosed in childhood, and say which classification you have applied.
Why this score exists
The working party was not trying to produce a phenotypic vocabulary. The report's title points toward an integrated clinical, molecular and serological classification, and the explicit expectation was that genetic and serological markers would be folded in as the evidence matured — the clinical axes were the part that could be agreed immediately. Twenty years on, that integration has not arrived in routine practice, and what remains in daily use is the layer the authors regarded as provisional scaffolding. The three Crohn's axes have held up well because each was chosen for a clinical reason: age at diagnosis because paediatric-onset disease behaves differently, location because it determines drug delivery and cancer risk, and behaviour because it is the axis that progresses and forces surgery. The authors were more openly uncertain about the ulcerative colitis severity axis, and the subsequent history has borne that out — validated activity indices have largely displaced S in practice while E remains the standard.
About the creator
First author, Montreal Working Party report
First author of the 2005 working party report that defined the Montreal classification.
Co-author; author of the 2006 consensus commentary
Co-authored the working party report and wrote the subsequent account of its controversies and implications.
First author, Paris paediatric modification
Led the paediatric consensus that adapted Montreal for children as the Paris classification.
Limitations
- Purely descriptive, with no predictive output of its own — its usefulness comes from the prognostic associations of individual descriptors, not from the classification as a whole.
- The ulcerative colitis severity axis is the least secure element and was identified as contentious from the outset; validated activity indices have largely replaced it in practice.
- Behaviour in Crohn's disease is hierarchical, so a patient with both stricturing and penetrating disease is recorded as B3 and the stricture disappears from the label.
- No descriptor for disease duration or for cumulative bowel damage, despite both being central to modern Crohn's prognostication — the Lémann index exists to fill that gap.
- The molecular and serological layers the working party intended have never been incorporated, so the classification in use is only part of what was designed.
- Assignment of L4 as a modifier is inconsistently applied in practice, which undermines comparability between cohorts — the axis Montreal specifically improved is the one most variably recorded.
- Extent in ulcerative colitis depends on the completeness of the endoscopic assessment, so an incomplete colonoscopy can produce a permanent under-classification with surveillance consequences.
- Not designed for paediatric practice, and the Paris modification exists because the single A1 category and the absence of a growth descriptor made it unusable there.
If you are the patient
The Montreal classification is a shorthand doctors use to write down what type of inflammatory bowel disease you have and where in the bowel it affects. It is not a score and it does not say how bad things are today — there is a separate set of measures for that. For Crohn's disease it records three things: how old you were when diagnosed, which parts of the bowel are involved, and whether the disease has caused narrowing or has tunnelled through the bowel wall. A small 'p' is added if there is disease around the back passage. For ulcerative colitis it records how far up the colon the inflammation reaches, and a rough sense of current severity. You may see it written as something like A2L3B2p or E3S2. Two parts of it directly affect your care. In Crohn's disease, narrowing or tunnelling usually means stronger treatments are considered earlier and a surgeon is involved in planning, because those changes do not respond to anti-inflammatory drugs in the way pure inflammation does. In ulcerative colitis, if the inflammation reaches beyond the lower part of the colon, you will be offered regular surveillance colonoscopies starting about eight years after your symptoms began, to look for early changes. One thing worth knowing: the recorded extent is the furthest the inflammation has ever reached, not what it looks like now — so it stays on your record even when you are well, because that is what your surveillance is based on.
Frequently asked questions
What is the Montreal classification of IBD?#
A consensus system for recording the phenotype of inflammatory bowel disease. Crohn's disease is coded by age at diagnosis (A1 ≤ 16, A2 17–40, A3 > 40), location (L1 ileal, L2 colonic, L3 ileocolonic, L4 isolated upper gastrointestinal) and behaviour (B1 non-stricturing non-penetrating, B2 stricturing, B3 penetrating), with a lower-case p for perianal disease. Ulcerative colitis is coded by extent (E1 proctitis, E2 left-sided, E3 extensive) and severity (S0–S3).
What do L1, L2, L3 and L4 mean in Crohn's disease?#
L1 is ileal disease, L2 is colonic, L3 is ileocolonic, and L4 is isolated upper gastrointestinal disease. L4 can also be added as a modifier to L1, L2 or L3 when upper gastrointestinal disease coexists with more distal disease — so L3+L4 is valid. That dual role was one of Montreal's changes from the Vienna classification, where L4 was mutually exclusive with the others.
What is the difference between B2 and B3 Crohn's disease?#
B2 is stricturing disease — fixed luminal narrowing — and B3 is penetrating disease, meaning fistulae or perforation. The categories are hierarchical, so a patient with both a stricture and a fistula is recorded as B3. Both indicate structural damage and both argue for earlier biologic therapy and early surgical involvement, but they need different assessments: a stricture requires imaging to distinguish inflammatory from fibrotic narrowing, while penetrating disease requires exclusion of an abscess before immunosuppression.
What do E1, E2 and E3 mean in ulcerative colitis?#
E1 is ulcerative proctitis, limited to the rectum. E2 is left-sided or distal disease, extending up to but not beyond the splenic flexure. E3 is extensive disease or pancolitis, extending proximal to the splenic flexure. Extent determines the route of therapy — topical for E1, oral or systemic for E2 and E3 — and E2 and E3 carry an obligation to begin colorectal cancer surveillance approximately eight years from symptom onset.
Does the Montreal classification change over time?#
Only the behaviour axis should be expected to. Age at diagnosis is fixed permanently, location is largely stable, and ulcerative colitis extent is recorded as the maximum ever documented so it does not decrease. Behaviour progresses: patients move from B1 to B2 or B3 as strictures and fistulae develop, and that transition is prognostically meaningful, which is why the descriptor should be dated in the record.
Is the Montreal classification a severity score?#
No. It produces a label, not a number, and there is nothing to total or threshold. The one axis that describes current state — S0 to S3 for ulcerative colitis — is the weakest part of the classification and has largely been replaced in practice by validated activity indices such as the Mayo score, SCCAI, UCEIS or Truelove and Witts.
What does the 'p' mean in a Montreal classification?#
Perianal disease, appended as a lower-case p to the behaviour category — for example B1p or B3p. It is independent of behaviour, so perianal fistulising disease does not make a patient B3. Montreal separated the two deliberately, because the Vienna classification folded perianal disease into penetrating behaviour and could not represent a patient with perianal disease and no internal penetrating complication.
How does Montreal differ from the Vienna classification?#
Four changes. Montreal added a paediatric age category (A1 ≤ 16, A2 17–40, A3 > 40) where Vienna split only at 40; it allowed L4 to be used as a modifier alongside L1–L3 rather than exclusively; it separated perianal disease into an independent p modifier; and it extended the classification to ulcerative colitis, which Vienna did not cover.
Should the Paris or Montreal classification be used in children?#
Paris. It subdivides A1 into A1a (under 10) and A1b (10 to under 17), adds E4 for pancolitis, allows stricturing and penetrating behaviour to be recorded together rather than hierarchically, and adds a growth-failure descriptor that Montreal cannot express. Those are exactly the descriptors paediatric management depends on.