What it does not check
Backdating, payment mode and sampling, and how to read a population that may be incomplete.
Read this before you rely on a report. Every scrutiny report prints these limitations on its own face, in the same words, so that nobody reading it has to guess what was covered.
Three things this version does not test at all
- Backdated entries
Telling a backdated entry from a late one needs the date it was TYPED, and that is not among the fields the connector asks Tally for. The alteration id it does fetch gives order, not time. Tally's own audit trail is the sound source and reading it is a future piece of work.
- How a payment was made
Whether a payment went by account payee cheque, bearer cheque or transfer is in the voucher's bank allocation, which is not fetched. So the cash tests read the cash ledgers and nothing else, and section 40A(3)'s account-payee question is not answered here.
- Sampling
No sample is drawn, for SA 530 or anything else. A sample is a claim about a population, and the completeness of the population cannot yet be proven — see below.
What may be missing from the population
Tally truncates a month that returns about five thousand vouchers or more, and the connector can only warn that it happened. When it does, the months affected are named at the top of your findings and again in the report.
A clean list over a truncated month is worse than no list. If you see that warning, narrow the run to a shorter period and run it again.
Where a test could not run at all
A test with nothing to work against reports that it did not run, with the reason. That is deliberately different from reporting that it found nothing, because the two look identical and mean opposite things.
The commonest case is the related-party test before anyone has recorded a related party.
Where cash was identified by name
Cash and loan ledgers are normally found by where they sit in Tally's group tree, which is why a till called "Counter Float" is caught and a debtor called "Cash Traders" is not. If a run has no ledger list — an older fetch, or one whose masters failed — the names are matched instead, and the report says so. Treat those cash findings as indicative and fetch again.
Section 269ST is only partly covered
The aggregation tested is per person per day. The section also catches a single transaction, and transactions relating to one event or occasion. Neither is tested, and every 269ST finding says so.
Last updated 18 Sept 2026
Still stuck? Ask us