GSTR-1 vs GSTR-3B for Credit Assessment: What a Mismatch Really Tells an Underwriter

GSTR-1 reports what the borrower invoiced. GSTR-3B reports what tax the borrower actually paid. A gap between the two is normal — credit notes, exports, advances and reverse charge all sit in different tables. A gap that is large, one-directional, persistent and unexplained is a liquidity signal, not a clerical one.

Key facts

• The comparison is now built into the GST system itself. Rule 88C of the CGST Rules requires the portal to intimate a taxpayer in FORM GST DRC-01B where "the tax payable by a registered person, in accordance with the statement of outward supplies furnished by him in FORM GSTR-1 or using the Invoice Furnishing Facility in respect of a tax period, exceeds the amount of tax payable by such person in accordance with the return for that period furnished by him in FORM GSTR-3B", and gives the taxpayer seven days to pay or explain (CBIC, *Notification No. 26/2022-Central Tax*, 26 December 2022).
• Ignoring the intimation blocks the next return. "If a taxpayer doesn't file response to Form GST DRC-01B for previous tax period, then for the subsequent tax period, they will not be able to file their Form GSTR-1/IFF" (GST portal, *FAQs > Return Compliance in Form DRC-01B*).
• YuSight extracts at 95.2% accuracy, validated against a manual benchmark — which matters because this comparison is built from 24 monthly returns per GSTIN, and a single mis-keyed table turns a clean file into a false positive.
• The dataset is near-universal. "The number of GST taxpayers increased from 66.5 lakh in 2017 to 1.65 crore as on May 2026", against FY 2025-26 gross collections of about ₹22.27 lakh crore (PIB, *Nine Years of GST*, 30 June 2026).
• Neither return can be revised. "Form GSTR-3B once filed cannot be revised" (GST portal, *FAQs > Form GSTR-3B*); GSTR-1 is corrected only through GSTR-1A or a later period's amendment tables.

What does each return actually report, and when?

The two returns are not two views of the same number. They are filed on different dates, built from different tables, and answer different questions.

Sources: GST portal, *FAQs > Form GSTR-1* and *FAQs > Form GSTR-1A*.

Two consequences for an underwriter. First, GSTR-1 is the closer proxy for turnover, because it is invoice level and customer level; GSTR-3B is the closer proxy for cash discipline, because filing it costs money. Second, the sequence matters: GSTR-1 is filed nine days before GSTR-3B, and GSTR-1A closes the moment GSTR-3B is filed. A borrower who is short of cash therefore has a nine-day window in which the sale is already on public record and the tax is not yet paid. That window is where the signal lives.

For the wider framework these two returns sit inside — GSTR-2A, 2B, 9 and 9C — start with our GST return analysis framework for Indian credit teams.

Why are the two numbers never supposed to be identical?

Before you treat a gap as a finding, eliminate the structural reasons the two returns must differ. Most "mismatches" an analyst reports are the analyst comparing the wrong tables.

Table 3.1 line references per GST portal, *FAQs > Form GSTR-3B*.

Which gaps are not legitimate?

Once the eight structural causes above are stripped out, only a short list remains, and each of them is a credit issue:

• Tax declared but not discharged. The sale is admitted invoice by invoice; the money to pay the tax on it was not there. This is the single most common finding, and it is a working-capital diagnosis.
• Deliberate under-declaration in GSTR-3B. The borrower files GSTR-1 correctly — because customers need the ITC — and files a lower 3B to defer cash outflow, intending to true up "next quarter". The true-up rarely comes.
• Invoices raised without a supply. GSTR-1 inflated to support a turnover-based limit or to accommodate a customer's ITC. Look for the mirror image: e-way bills that do not move, and customer concentration that appears suddenly.
• Circular billing inside a group. Sales inflated in both directions between related GSTINs, with no external cash. GSTR-1 rises, GSTR-3B rises, and bank credits do not.

What decision tree should an underwriter follow?

Work in this order. Stop as soon as the gap is explained.

1. Align the tables before you compare anything. Sum GSTR-1 Tables 4A, 5A, 6A, 7 and 9A/9B/9C; sum GSTR-3B Tables 3.1(a), (b), (c) and (e). Never compare a GSTR-1 total against 3.1(a) alone.
2. Check the filing frequency. If the borrower is on QRMP, compare quarters, not months.
3. Check the sign. GSTR-1 higher than GSTR-3B is the cash-stress direction. GSTR-3B higher is usually reverse charge, advances or an error in your own extraction.
4. Size it against turnover. Under 1% of annual outward supplies is noise. Between 1% and 5%, bridge it. Above 5%, treat the turnover as unproven until the bridge closes.
5. Test persistence. One month is a cut-off. Every month, in the same direction, growing, is a trend.
6. Read the filing dates alongside the values. A widening value gap that arrives with lengthening GSTR-3B delays is a liquidity story with two independent witnesses.
7. Ask for the DRC-01B. If the system issued an intimation under Rule 88C, ask for the intimation and the reply. A borrower who says there was no mismatch and cannot produce a clean Return Compliance tab has told you something.
8. Corroborate outside GST. Take the residual to bank credits and to the ITR — the three-way triangulation step. Where the ITR is the anchor, read it schedule by schedule using ITR analysis for loan underwriting.

Worked example one: a ₹1.84 crore gap that resolves innocently

All figures below are illustrative and constructed to demonstrate the method. They are not client data.

Anand Polymers Private Limited, a single-GSTIN moulder in Vadodara, FY 2025-26. The analyst's first pass:

The bridge:

The remaining ₹16,00,000 traced to a single March 2026 invoice to a state transport undertaking, reported in the March GSTR-1 with output tax discharged through the April 2026 GSTR-3B. Pulling the April 2026 return confirmed it. At 18% that is ₹2,88,000 of tax — nowhere near any escalation threshold, and it reverses in the following year.

Verdict: clean. Record the bridge in the CAM and move on.

Worked example two: a gap that does not resolve

Illustrative. Suryodaya Fabrics LLP, Surat, FY 2025-26. Same method, aligned tables from the start.

The monthly view is what settles it. Seven of the twelve months (₹ lakh):

These seven months carry ₹3.72 crore of the ₹3.96 crore residual. The remaining five months contribute ₹24 lakh between them.

Read it line by line. Sales are rising. The gap appears in October, never reverses, and grows. The GSTR-3B filing date lengthens in step with the gap and then shortens in February — the classic pattern of a borrower who catches up just before a facility review. Nothing here is a cut-off difference; cut-off differences do not compound.

What it costs the file:

Three consequences follow. The arrear is a creditor you were not counting, and it should come out of current assets before you compute drawing power. A Rule 88C intimation is almost certainly outstanding, and non-response blocks the next GSTR-1 — which stops the borrower's customers claiming ITC and puts the order book at risk. And the borrower's own explanation, whatever it is, is now testable: if there was cash, why was the tax not paid?

Verdict: the turnover is real, the liquidity is not. This is not a decline on integrity grounds. It is a resizing, a covenant on statutory-dues clearance, and a monthly GSTR-3B filing condition.

How does YuSight handle this comparison?

GST returns reach a credit file as portal PDFs, JSON exports, screenshots of the Returns dashboard and photographs of a filed acknowledgement. YuSight's Document Intelligence classifies each one, reads the GSTIN and tax period off the face of the document, and maps it to the right borrower entity — the difference that matters when a group sends three years of returns for nine GSTINs in one folder with no naming convention.

Financial Spreading then builds the aligned GSTR-1 to GSTR-3B series table by table, the monthly gap and the filing-delay calendar, and carries every figure into the CAM with a citation back to the page it came from — alongside the ITR and Form 26AS series described in ITR and Form 26AS spreading for MSME lending. Extraction runs at 95.2% accuracy against a manual benchmark, and every number stays analyst-editable.

FAQ

What is the difference between GSTR-1 and GSTR-3B?

GSTR-1 is an invoice-by-invoice statement of what the borrower sold. GSTR-3B is a summary return where the tax on those sales is actually paid. One is a claim about sales, the other is a payment.

Why would GSTR-1 and GSTR-3B differ?

Usually for a boring reason. Credit notes, exports, advances and reverse charge all sit in different tables, so if you sum the wrong ones you manufacture a gap that was never there. Align the tables first, then look at what is left.

Is a GSTR mismatch a decline trigger?

On its own, no. A gap under about 1% of turnover is noise. What changes the decision is a gap that is large, always in the same direction, present every month and still unexplained after you have bridged it.

What does it mean when GSTR-1 is higher than GSTR-3B?

Most often that the borrower has declared the sales and has not paid the tax on them. The sales are already on record invoice by invoice, so this is a cash-flow signal rather than a concealment one — and the unpaid tax is a senior creditor you should be counting.

What is Form DRC-01B and should I ask for it?

It is the intimation the GST system issues under Rule 88C when GSTR-1 liability exceeds GSTR-3B liability by more than the notified limit. Yes, ask for it. If one was issued, the borrower had seven days to pay or explain, and the reply tells you what they said when the tax department asked the same question you are asking.

Can the borrower simply amend GSTR-1 to make the gap go away?

Only within a narrow window. GSTR-1A closes the moment GSTR-3B for that period is filed, and it cannot change the recipient's GSTIN at all. After that, corrections have to be made in a later period's amendment tables, where you can still see them.

How many months of returns do I need for this to be meaningful?

Twenty-four. Twelve months shows you the shape of a year but not whether this year is worse than the last one, and a gap that only started nine months ago looks like a permanent feature in a twelve-month window.

Does this work for a QRMP borrower?

Yes, but only at quarterly granularity. A QRMP filer pays tax monthly through PMT-06 and files GSTR-3B quarterly, so a month-on-month comparison of the two returns will show gaps that mean nothing at all.

Where should I take the residual once I have bridged it?

To the bank statements and the ITR. A GST residual that also appears as missing bank credits is a very different finding from one that does not, which is why the reconciliation is properly a three-way exercise rather than a two-way one.

Key takeaways

• GSTR-1 measures what was invoiced; GSTR-3B measures what was paid. They are supposed to differ.
• Eight structural causes — credit notes, exports, exempt supplies, reverse charge, advances, prior-year amendments, QRMP and year-end cut-off — explain most apparent mismatches. Eliminate them before you escalate anything.
• Compare aligned tables, not headline totals. Most false positives come from summing GSTR-1 against Table 3.1(a) alone.
• Size the residual against turnover, then test whether it is persistent and one-directional. Direction and persistence carry more information than magnitude.
• Read the value gap and the GSTR-3B filing dates on the same page. When both deteriorate together, you have a liquidity finding with two witnesses.
• An unpaid GST liability is a statutory creditor ranking ahead of you. Deduct it before you compute drawing power, not after.

Upload a messy document set and see it classified — three years, nine GSTINs, no naming convention. Book a live demo.