COMLEX and USMLE: what changes in the reasoning tested
Table of contents
Two multiple-choice booklets. One exam day. Hundreds of best-answer items. The difference is not just in the logo: COMLEX-USA Level 2-CE (NBOME) and USMLE Step 2 CK (FSMB/NBME) demand the same decision-making muscle under a clinical vignette, but with distinct blueprints, blocks and emphases. In 2026 both systems also change the shape of the day: Level 2-CE moves to 320 items in eight sections of 40; Step 2 CK, starting May 7, 2026, uses sixteen 30-minute blocks.
NEXO is a daily medical diagnosis game: one clinical case a day, six progressive clues, and ICD-10 proximity feedback across five levels.
This text compares what each exam asks of clinical reasoning, how the 2026 format changes the rhythm of the day, where Osteopathic Principles and Practice (OPP) fits into COMLEX and how to train without mixing question banks as if they were the same instrument. Anyone who has read the series on USMLE Step 2 CK, ENARM, Examen MIR or ENARE recognizes the family: application of knowledge in a written case. The rule specific to this installment is the side-by-side comparison, with NBOME and USMLE sources.
Key points
- COMLEX-USA Level 2-CE is an assessment of clinical sciences and osteopathic competencies for entry into supervised training; starting in June 2026, 320 best-answer items in two four-hour sessions, eight sections of 40 (NBOME Examination Format; 2026 enhancements).
- USMLE Step 2 CK assesses application of clinical science essential to supervised care; until May 6, 2026, eight 60-minute blocks (up to 318 items); starting May 7, 2026, sixteen 30-minute blocks (up to 20 items per block), still in a nine-hour session (USMLE Step 2 CK; May 2026 announcement).
- In the current COMLEX blueprint, OPP/OMT weighs about ten percent on Level 2-CE (series minimum ten percent); confirm the table on the NBOME site (blueprint).
- Both sit on the "knows how" rung of Miller: written vignette, not OSCE. OPP and osteopathic language are the structural difference of COMLEX; Step 2 CK does not test OMT as its own domain.
- Training by mixing question banks without looking at the blueprint wastes time: the decision node may be similar; the distribution of emphases and the rhythm of the blocks are not.
Two exams, one family of decision
USMLE Step 2 CK is the clinical knowledge exam of the USMLE series, used by MD graduates (and by many DOs who also take it) on the path to licensure and in US residency ranking. COMLEX-USA Level 2-CE is the corresponding exam of the NBOME series for osteopathic medical students and graduates: it integrates clinical sciences, osteopathic principles and supervised care competencies (NBOME Level 2-CE; USMLE Step 2 CK).
On Miller's pyramid, both sit on the second rung: knowing how to apply knowledge in written cases (Miller, 1990). They do not replace the skills stage with standardized patients (now redesigned in each ecosystem). They are siblings of MIR, ENARM and NEET-PG: decision under a stem, plausible distractors, best answer.
Who organizes what
FSMB and NBME administer the USMLE; NBOME administers COMLEX-USA. They are distinct series, with their own BOI, testing windows, scoring and policies. US residency programs may accept COMLEX, USMLE or both: the program's policy rules, not the forum. If your target lists only one, read the program's site and the current BOI before setting your calendar.
What "passing" communicates
Passing Level 2-CE indicates competence in clinical sciences and osteopathic principles to enter supervised training and keep learning (NBOME description). Passing Step 2 CK indicates application of clinical science essential to care under supervision, with emphasis also on health promotion and prevention (USMLE). The official sentences are not interchangeable: one carries OPP explicitly; the other does not.
2026 formats side by side
COMLEX Level 2-CE (starting June 2026)
One-day administration, computer, timed environment. Two four-hour sessions (A and B); time not used in session A does not carry over to B. Eight sections of 40 questions (320 in total), best answer, with possible audiovisual elements. Sixty minutes of pooled break in the scheduled windows. Pre-exam tutorial of about 16 minutes. Items can be reviewed within the section; when you move on, the previous section closes (Examination Format; enhancements).
Up to the previous cycle, the BOI still described 352 items (sections of 44). The reduction to 320 aligns the booklet with the blueprint; the total test time remains eight hours. Train in the new envelope if your date is June 2026 or later.
USMLE Step 2 CK (cutoff on May 7, 2026)
Before May 7, 2026: one day, eight blocks of 60 minutes, up to 40 items per block, total not exceeding 318, nine-hour session, at least 45 minutes of break and optional 15-minute tutorial.
From May 7, 2026: same nine-hour day and item ceiling; sixteen blocks of 30 minutes, up to 20 items per block; at least 55 minutes of break and optional 5-minute tutorial. The official announcement makes clear that the content and the length of the day do not change; what changes is the delivery software and the block size (USMLE; software May 2026).
Comparison table
| Item | COMLEX Level 2-CE (Jun/2026+) | USMLE Step 2 CK (May 7/2026+) |
|---|---|---|
| Organizer | NBOME | FSMB / NBME |
| Items (ceiling / target) | 320 best-answer MCQ | up to 318 items |
| Day envelope | 2×4 h of testing + breaks | 9 h session |
| Blocks / sections | 8×40 | 16×30 min (≤20 items) |
| Break | 60 min pooled | ≥55 min |
| Own domain | OPP/OMT in the blueprint (about ten percent of L2-CE) | no OMT as its own domain |
| Miller's step | knows how (vignette) | knows how (vignette) |
Where clinical reasoning shows up
In both exams the stem describes the patient, the context and the findings; the options offer diagnoses, next steps or interpretations. The typical mistake is not "not knowing the name of the disease": it is leaving the stem too early, anchoring on the first hypothesis or choosing the "almost right" option that ignores the decisive data. This is the same logic of key features discussed in the assessment literature (Page and Bordage, 1995; Bordage and Page, 2018).
Block pace and cognitive load
Shorter blocks (Step 2 CK after May) force a mental restart more often: open the stem, close it, next item, and do not go back to the previous block. Sections of 40 in COMLEX call for stamina within the section and the discipline of not "saving review" for a section that has already closed. In both, time not used in one unit does not fix the next one the way an eternally open block would.
Sequences and multimedia
Step 2 CK includes sequential items, scientific abstracts and audio or video findings in the practice material (exam content). COMLEX also allows audiovisual elements. When the exam presents an image or sound, the reasoning requires integrating the finding into the stem, not treating the stimulus as decoration.
Cross-cutting clinical work
In both blueprints clinical work runs across systems and contexts (outpatient, emergency, ward). Training that only memorizes lists by specialty without training the decision node fails in the same place in COMLEX and in USMLE: a seductive distractor, a decisive data point ignored.
OPP and what USMLE does not test
Osteopathic Principles and Practice / OMT is not a "cultural appendix" in COMLEX: it enters the blueprint as a competency domain. In the current table cited by NBOME, OPP/OMT corresponds to about ten percent of Level 2-CE (minimum of ten percent of the series); the future Blueprint 2.0 raises the series minimum, with its own schedule on the NBOME site (introduction to the blueprint). Always confirm the table for the cycle in which you take the exam.
In practice, this changes the reasoning tested in three ways:
- Items in which the best answer is an OMT maneuver, indication or contraindication.
- Items in which osteopathic language (structure-function, viscerosomatic, barrier) selects the correct option among plausible "general medical" distractors.
- General clinical items in which OPP appears as a safety or reasoning filter, not as a curiosity.
Step 2 CK has no equivalent OMT domain. A DO who takes both exams trains two envelopes: the USMLE bank covers the shared clinical content; the COMLEX bank requires the OPP slice. An MD who only takes the USMLE does not need to invent OMT; they need vignette reasoning in the Step format.
Who takes what in practice
The osteopathic school student (DO) has the COMLEX as their own series. Many US residency programs accept COMLEX Level 2-CE in place of Step 2 CK; others require USMLE or accept both. The decision to take Step 2 CK in addition to COMLEX is strategic: it widens match options, and it requires a second envelope of blocks and a second error bank. It is not mandatory under any single federal law; it is a reading of program and specialty requirements.
The MD student typically takes the USMLE. They do not need to invent OPP. If they ever come across COMLEX material out of curiosity, the gain is in the shared clinical content, not in the OMT slice. Mixing banks "because the stems look alike" without checking the blueprint is the most expensive mistake in this comparison.
For those who come from another country and compare with ENARM, MIR or ENARE, the useful parallel is: all of them demand a decision under a vignette; only the COMLEX ties OPP in as a domain sampled explicitly in the NBOME blueprint.
What does not change between the two
- Best answer, not free-form essay.
- Stem with enough data and plausible distractors.
- No going back to sections or blocks already closed (the discipline of closing the item).
- The need for a full day of stamina.
- The importance of reading the BOI or the official page for your cycle, not a group summary.
These constants explain why a good reasoning habit (read to the end, find the crux, eliminate) transfers. The differences in format and in OPP explain why the study plan cannot be a single untagged Anki deck.
What the studies say
The literature on clinical reasoning and medical assessment explains why two national MCQs can correlate and still differ in emphasis. Eva describes the combination of analytic and non-analytic processes in diagnosis (Eva, 2005). Norman and Eva link diagnostic error to failures of these processes, not only to a "lack of facts" (Norman and Eva, 2010). Van der Vleuten and Schuwirth argue that professional competence is best assessed with broad sampling and combined methods (van der Vleuten and Schuwirth, 2005).
For the candidate, the translation is prosaic:
- Broad sampling (hundreds of items) reduces the bad luck of an isolated topic.
- OPP on the COMLEX is deliberate sampling of a domain that Step 2 CK does not mirror.
- A change in block size alters fatigue and strategy, not the definition of competence.
No article replaces the BOI for your cycle. Use the literature to understand the design; use NBOME and USMLE for the number of items and the cutoff date.
Published comparisons between COMLEX and USMLE performance exist in the educational literature, but correlation numbers from an article do not replace the requirement of the program you want to match into. Use articles to understand that the exams measure related but not identical constructs; use the program website for the application rule. Eva and Norman remind us that diagnostic error mixes knowledge failure and process failure: training only facts or only "exam tips" leaves one side uncovered (Eva, 2005; Norman and Eva, 2010).
In the practice of the error notebook, this becomes three columns: missing fact, process (anchor, leaving the stem early), and OPP domain. Without the third column, the DO candidate underestimates the slice that the blueprint reserves on purpose.
How to train without mixing the banks
Stem discipline
Read the stem to the end before looking at the options. Mark the data point that changes the decision (time, severity, contraindication, discordant finding). Only then eliminate. This habit serves both exams and also Revalida when the logic is a decision under a case.
Mock exam in the right envelope
- COMLEX Jun/2026+: blocks or sections of 40, two long sessions, practice closing a section without going back.
- Step 2 CK from 7 May/2026: 30-minute blocks, practice frequent restarts and the new software (official tutorial).
- Step 2 CK before that date: keep eight 60-minute blocks in the mock exam.
OPP with real weight
If you take COMLEX, set aside explicit weekly time for OMT/OPP in the blueprint, not "if there is time left". If you take only USMLE, do not divert hours to OMT at the expense of shared clinical content.
Example of a shared node (fictional, for method)
A stem describes chest pain in a young adult after exertion, with a discordant finding on examination. The options mix embolism, pericarditis, costochondritis and anxiety. On Step 2 CK and on the general clinical Level 2-CE, the move is the same: which data point decides? Timing, friction rub, saturation, D-dimer, ECG. The "almost right" option that ignores the rub or the post-viral context is the classic distractor.
Now add an OMT option or an osteopathic formulation of thoracic somatic dysfunction with a technique indication. On COMLEX, that line can be the best answer or a distractor that only looks "home-grown". On USMLE, that line simply does not appear as its own domain. Drilling both banks with the same "chest pain" flashcard list without marking OPP items hides the difference.
Managing exam day
On COMLEX with two four-hour sessions, the operational mistake is spending session A on endless review within the section and reaching B with no cognitive margin. On Step 2 CK with sixteen short blocks, the mistake is treating each block as a sprint with no break plan: the pooled break or the minimum break only helps those who schedule it. On both, finishing the tutorial early can lengthen the break; this is in the official pages and is worth rehearsing in the software.
When to stop mixing
Stop mixing when:
- Your OPP errors fall into their own category and you still count them as "general clinical".
- Your time per item in the new envelope (40 per section or 30 minutes) is off the pace of the old mock exam.
- You change your date and cross the May (USMLE) or June (COMLEX) cutoff without redesigning the mock exam.
Error log by type of decision
Separate reading, knowledge, OPP and timing errors. Mixing everything into one list by "clinical topic" hides the pattern the exam actually punishes.
Preparation checklist
- [ ] I confirmed the cutoff date for my Step 2 CK (before/after May 7, 2026)
- [ ] I confirmed whether my Level 2-CE is already 320 items (Jun/2026+)
- [ ] I read the NBOME BOI and the official Step 2 CK page for this cycle
- [ ] I set aside an OPP slice if I take COMLEX
- [ ] I did at least one mock exam in the block envelope I will face
- [ ] Error log by type of decision, not only by organ
- [ ] I checked the COMLEX/USMLE policy of the target programs
- [ ] Tutorial for the new software (if date ≥ cutoff)
How NEXO helps you train
In researcher mode, you explore hypotheses more calmly; in closer, you practice closing under time pressure; curator and the creators area help you review what you have already seen. Usage limits and the day's economy are at /economy. There is no claim here that NEXO raises scores on COMLEX or USMLE: what the product offers is deliberate practice of reasoning under a case.
If the typical error is "I left the stem too early" or "I anchored on the first hypothesis", a well-played case forces the same discipline as the right item. Useful links: /researcher, /closer, /curator, /creators. For clue design and differential, see differential diagnosis with clues and case-based learning.
Frequently asked questions
Are COMLEX and USMLE the same exam with different names?
No. They are distinct series (NBOME vs FSMB/NBME), with their own blueprints, policies and emphases. The vignette clinical content looks similar; OPP and administration rules do not.
Do I need both if I am a DO?
It depends on the target programs and your strategy. Many DOs take COMLEX; some add USMLE. Read each program's policy and the current BOI.
What changes on Step 2 CK in May 2026?
Delivery software, sixteen 30-minute blocks (up to 20 items), a longer minimum break and a shorter tutorial. Content and the length of the day remain in the design announced by USMLE.
Does Level 2-CE still have 352 questions?
The previous BOI described 352. For administrations from June 2026 onward, the published format is 320 items in eight sections of 40. Confirm on the NBOME page for your cycle.
How much OPP is on Level 2-CE?
In the current blueprint table cited by NBOME, about ten percent on Level 2-CE (series minimum ten percent). Check the updated table on the official site.
Do shorter blocks make the exam easier?
Not automatically. They change the pace and the review strategy within the block. The sampling of competence remains broad.
How does NEXO fit into preparation?
As practice in reasoning under clues and hypotheses, complementary to the official banks. No promise of a score; with a habit of deciding.
Final checklist
- [ ] I know which block envelope applies on my date
- [ ] I set aside OPP if I am taking COMLEX
- [ ] I did not treat COMLEX and USMLE as the same bank
- [ ] Mock exam in the official format of the cycle
- [ ] Errors classified by type of decision
- [ ] Policies of the target programs checked
- [ ] Software tutorial (new USMLE, if applicable)
- [ ] BOI/NBOME and the USMLE page open, not a forum summary
Play the case of the day
Open the case of the day and practice putting the clues together before you commit. Then go back to the COMLEX or USMLE bank with the same discipline: full stem, decisive data, best answer.
References
- NBOME. COMLEX-USA Examination Format (Level 1 and Level 2-CE; 320 items, 2×4 h). [nbome.org](https://www.nbome.org/assessments/comlex-usa/examination-format/).
- NBOME. Enhancements to COMLEX-USA Exams in 2026, 2027 (320 items; Level 2-CE from June 2026). [nbome.org](https://www.nbome.org/news/enhancements-to-comlex-usa-exams-in-2026-2027/).
- NBOME. Description of Examinations (BOI): Level 2-CE format and passing. [nbome.org](https://www.nbome.org/assessments/comlex-usa/bulletin-of-information/description-of-examinations/).
- NBOME. COMLEX-USA Level 2-CE (exam page). [nbome.org](https://www.nbome.org/assessments/comlex-usa/comlex-usa-level-2-ce/).
- NBOME. COMLEX-USA Blueprint: introduction (OPP/OMT and weights). [nbome.org](https://www.nbome.org/assessments/comlex-usa/comlex-usa-blueprint/introduction/).
- USMLE. Step 2 CK (formats before and from May 7, 2026). [usmle.org](https://www.usmle.org/step-exams/step-2-ck).
- USMLE. Test Delivery Software Updates for Step 2 CK and Step 1 Coming in May 2026. [usmle.org](https://www.usmle.org/test-delivery-software-updates-step-2-ck-and-step-1-coming-may-2026).
- USMLE. Step 2 CK Exam Content. [usmle.org](https://www.usmle.org/step-exams/step-2-ck/step-2-ck-exam-content).
- Miller GE. The assessment of clinical skills/competence/performance. Acad Med. 1990. [PubMed](https://pubmed.ncbi.nlm.nih.gov/2400509/).
- Page G, Bordage G. The Medical Council of Canada's key features project. Acad Med. 1995. [PubMed](https://pubmed.ncbi.nlm.nih.gov/7748379/).
- Bordage G, Page G. The key-features approach. Med Educ. 2018. [PubMed](https://pubmed.ncbi.nlm.nih.gov/29777464/).
- Eva KW. What every teacher needs to know about clinical reasoning. Med Educ. 2005. [PubMed](https://pubmed.ncbi.nlm.nih.gov/15813700/).
- Norman GR, Eva KW. Diagnostic error and clinical reasoning. Med Educ. 2010. [PubMed](https://pubmed.ncbi.nlm.nih.gov/20051831/).
- van der Vleuten CPM, Schuwirth LWT. Assessing professional competence. Med Educ. 2005. [PubMed](https://pubmed.ncbi.nlm.nih.gov/16147794/).