Methodology: how every number on this site is produced

The source

Every May the European Commission's Directorate-General for Migration and Home Affairs publishes an Excel workbook titled "Visa statistics for consulates" for the previous calendar year. Its sheet "Data for consulates" has one row for every consulate of every Schengen state anywhere in the world, with these columns: Schengen State; country where the consulate is located; consulate (city); airport transit visas (ATVs) applied for, issued, multiple, not issued and the ATV not-issued rate; uniform visas applied for; total uniform visas issued (including multiple-entry); multiple-entry uniform visas (MEVs) issued; share of MEVs; total LTVs issued; uniform visas not issued; not-issued rate for uniform visas; and totals combining ATVs and uniform visas.

We use the releases for 2009 through 2025, seventeen workbooks. The column layout of the per-consulate sheet is stable from 2013 onward. The totals sheets are not: the 2014 workbook's third-country totals sheet carries one column fewer than every other year, which is exactly what produced the 2014 error in the corrections log below. Every column is now resolved by its header text rather than its position, in both the India and the world extraction, and the build fails if the totals disagree with the sum of the consulate rows. The 2009 to 2012 files use different layouts (2011 and 2012 share one; 2010 and 2009 each differ) and were mapped column by column; the differences that matter are listed under "Older years" below.

What we keep

We keep every row where the "country where consulate is located" is India and the issuing state applied the Schengen rules in full that year: 697 consulate-year rows across the seventeen years, 28 Schengen states and 7 Indian cities (New Delhi, Mumbai, Chennai, Kolkata, Bengaluru, Goa, Puducherry). Rows for Bulgaria and Romania before 2024, Croatia before 2023 and Cyprus in any year are excluded because those states issued national visas then, not uniform Schengen visas. Liechtenstein has no consulate in India and is represented by Switzerland, so it has no rows; its page says so instead of inventing figures.

City names follow the Commission's spelling except that we display "Bangalore" as Bengaluru and "Pondichery" as Puducherry, and say so on those pages.

What we compute, and how

• Refusal rate = uniform visas not issued divided by uniform visas applied for, computed by us from the two counts for every row.

This is not the same denominator the Commission uses, and we should say so plainly. The Commission's own "not issued rate" divides refusals by applications decided (visas issued, plus limited-territorial-validity visas issued, plus visas not issued). We verified this against every row of the 2025 workbook: refused / (issued + LTV + refused) reproduces the Commission's published rate exactly. For India in 2025 the Commission's figure is 15.78%; ours is 15.70%.

We use applications lodged because that is the question a reader is asking ("of the people who applied, how many were refused"), and because it does not quietly drop applications that were withdrawn or still pending at year end. The cost of that choice is that our rate runs slightly below the Commission's. Measured across the 43 consulates in India above the ranking floor in 2025: mean difference 0.21 percentage points, largest 2.37 points (Slovakia in New Delhi, 21.09% ours against 23.45% theirs), and in one case ours is higher (Estonia in New Delhi, 30.07% against 29.69%). Rankings are barely affected: the top ten are identical either way. Both rates are published in the CSV on the data page, as refusal_rate_on_applied and refusal_rate_on_decided_ec_method, so you can use whichever you prefer and check us. - Multi-entry share = multiple-entry uniform visas issued divided by total uniform visas issued. - Country and city aggregates are sums of the consulate rows for that state or city in that year; the rate is then computed on the sums, never averaged across consulates. - All-India totals come from the workbook's "Totals - third country" sheet, and we check that the sum of our consulate rows matches it (it does, to the unit, for 2024 and 2025). - World totals are the sums of the "Schengen totals - applications" sheet, and India's rank is the position the Commission gives it in the third-country totals. - Year-on-year and long-run changes are simple differences in percentage points between the two years named on the page.

Older years (2009 to 2013)

• 2013 has the same columns as later years but no comparable third-country totals sheet; 2009 to 2012 have none either. For those five years the all-India total is the sum of the consulate rows, and India's world rank and the world refusal rate are shown as not available.
• 2011 and 2012 publish "C visas not issued" directly. 2009 and 2010 publish only "total A, B, C visas not issued"; we use that figure as the refusal count because airport-transit (A) and transit (B) visas were negligible at consulates in India, and we say so here rather than hide it.
• 2009 has no multiple-entry column, so multi-entry share is not available for that year.
• The Bengaluru consulates appear from 2011; earlier files list five or six Indian cities.
• Country names were normalised (for example "Czech Rep." to Czech Republic; state codes in 2010 mapped to names).

The ranking floor

Rankings, "best odds" labels and the #1/#last callouts include only states or consulates with at least 1,000 uniform visa applications in the year concerned. Without a floor a consulate with two applications and zero refusals would top every list. Below-floor consulates are still shown in tables and on their own pages, labelled with their application count, and are never given a rank.

What the figures do not cover

• Nationality. The Commission counts by where the application was lodged, not by the applicant's passport. "Applications lodged in India" are overwhelmingly Indian passport holders but include foreign residents of India. When we write "Indian applicants" as shorthand, this is what it means.
• National visas. Long-stay (type D) visas for study, work or family reunification are national decisions and are not in this data at all. Student-visa refusal rates for any country are a different dataset.
• Purpose. The workbook does not split tourism from business or family visits.
• Applications, not people. One person may lodge several applications in a year; refusals followed by successful reapplications count as one refusal and one issuance.
• Decisions pending or withdrawn. For some consulates, applied minus issued minus refused leaves a remainder: applications withdrawn, discontinued or still pending at year end. Our refusal rate keeps "applied" as the denominator, so that remainder sits in the denominator and is not counted as a refusal. See the note on the rate above for how this differs from the Commission's own published rate.
• Airport transit visas are excluded from headline figures (they are rare in India) and included as separate columns in the CSV.

Known anomalies you should read with care

• Some consulates report tiny volumes in a given year (for example the Netherlands consulate in Mumbai reported 2 applications in 2025, both issued). They appear on their pages with the count shown, and are excluded from rankings by the floor.
• Consulates appear and disappear across years as states open, close or centralise processing. A missing bar in a chart means "not in that year's file", not zero.
• 2020 and 2021 are pandemic years; volumes collapsed and rates behave unusually.
• Where a state centralises decision-making in one Indian city, the consulate that physically receives files may report few or no refusals while the deciding one reports many. We show the data as published and note it where visible.

Refresh cadence

When the Commission publishes the next release (normally in May), we re-run the same extraction, re-check the totals against the "Totals - third country" sheet, and rebuild every page. Every page shows the release range and the last-verified date in its footer.

Reuse

The Commission's documents may be reused under its reuse policy (Commission Decision 2011/833/EU, CC BY 4.0). Our derived CSV files are on the data page and may be reused with attribution to the European Commission for the source and to this site for the derivation.

Corrections we have had to make

Keeping this list public is part of the point of the site.

• 18 August 2026, the 2014 totals. An independent audit of the live site found that the year page for 2014, the homepage chart, the year hub, the by-year ranking, the totals CSV and two answer cards all reported zero refusals in India in 2014. The true figure is 36,519 refusals, a rate of 6.43%. Cause: the 2014 workbook's third-country totals sheet carries eight columns where every other year carries nine (it has no LTV column), and our extractor read that sheet by column position rather than by column name, so the refusal count was read out of the multiple-entry slot and the rate out of the refusal slot. Fixed by reading every value by its header text, by cross-checking the totals against the sum of our own consulate rows, and by making the build fail outright if those two independent paths ever disagree again. The consulate-level data for 2014, which the site's tables and charts are built from, was never affected and was correct throughout.
• 18 August 2026, the refusal-rate definition. This page previously claimed we compute the refusal rate "using the same definition" as the Commission. That was wrong: the Commission divides by applications decided, we divide by applications lodged. The section above now states the difference, quantifies it, and both rates are published in the CSV.
• 18 August 2026, a broken link in machine-readable data. The Dataset structured data on roughly half the pages pointed at a CSV filename that no longer existed after the series was extended back to 2009. Fixed to the live filename.

Corrections

If you can show a number that does not match the source workbook after the computations above, write to us via the contact page with the year, the consulate and the cell, and we will fix it and say so.

Printed from Schengen Odds (https://schengen-odds.pages.dev/methodology/). Sources are listed below; rules change, so check the linked official pages on the day you apply.