RRB ASM Exam Pattern 2026: CBT and CBAT

The RRB ASM exam pattern for 2026 differs from the NTPC pattern many candidates remember. Station Master is not part of CEN 06/2026, and third-party reports of a Railway Board letter dated 7 October 2026 describe a single-stage computer based test (CBT) followed by a computer based aptitude test (CBAT). This page lays out that reported pattern, compares it with the earlier two-stage one, and explains what it means for your preparation.

The official CEN for Station Master was not available on the RRB sites when this page was written, so all figures marked reported are provisional.

Table of Contents

• RRB ASM exam pattern at a glance

• CBT pattern

• Marking scheme and negative marking

• CBAT pattern

• Final merit calculation

• What the pattern means for you

• Earlier pattern for comparison

• Time management and exam-day strategy

RRB ASM exam pattern at a glance

CBT pattern

Two sites list the first section differently. One shows data analysis as a separate 30-mark block and another shows reading comprehension inside the logical section. These sources do not agree, so treat the three-section, 60-20-20 structure as the working pattern and wait for the CEN. The language options and whether sections are timed separately are also unconfirmed.

Marking scheme and negative marking

The earlier NTPC graduate CBTs gave one mark per question and deducted one-third of a mark for each wrong answer. For the Station Master CBT, the sources we reviewed do not state the negative marking. Until the CEN confirms it, practise as if the one-third deduction applies, because that is the safer habit. Unanswered questions were not penalised in the earlier pattern.

CBAT pattern

CBAT is a set of computer-based aptitude sub-tests. In the earlier NTPC cycle, coaching sites described five sub-tests: intelligence, selective attention, spatial scanning, information ordering and personality. They also reported a minimum T-score of 42 in each sub-test except the personality test, and no negative marking. This was not confirmed from an official page we could open, and the new process may change it.

• Shortlisting for CBAT: Reports say candidates up to about eight times the vacancies, in CBT merit order, will be called.

• Qualifying in CBAT: Reports say only candidates who qualify CBAT will be considered for combined merit.

Final merit calculation

Reports say the combined merit is 70 percent CBT and 30 percent CBAT, and that document verification and medical will be done for candidates up to the number of vacancies in that merit order. As a simple illustration, if a candidate's normalised CBT score is 60 percent and the CBAT score is 50 percent, the combined merit is 0.7 x 60 + 0.3 x 50 = 57 percent. This is only an example, as the actual normalisation method will be in the CEN.

What the pattern means for you

• Maths and science dominate: 60 of 100 marks come from one section, so a weak score here is hard to recover elsewhere.

• Reasoning is a safer scoring area: Logical and mental reasoning together give 40 marks, and these improve fast with practice.

• Time per question is tight: At 120 minutes for 100 questions, you have about 1.2 minutes each, a little more than the earlier pattern.

• CBAT counts for 30 percent: A strong CBT can be diluted by a poor aptitude test, so do not ignore it.

Earlier pattern for comparison

Both earlier CBTs had General Awareness, Mathematics and General Intelligence and Reasoning, and a one-third negative mark. This is the pattern used for the four posts in CEN 06/2026. It is not the reported Station Master pattern, so do not use it for planning timing.

Time management and exam-day strategy

1. Spend the first pass on questions you can solve in under a minute, across all sections.

2. Return to longer data and puzzle sets in the second pass.

3. Keep a short check for silly errors in the last ten minutes.

4. For CBAT, work steadily. Accuracy matters as much as speed in attention tasks.

Always verify the pattern on rrb.indianrailways.gov.in once the notification is out.