We Counted 999 Visa Interviews. Almost Everything You've Read About the Questions Is Wrong.
Search for visa interview prep and you will find the same article fifty times: a list of 50 questions, no sources, no dates, no indication that anyone was ever actually asked any of them. So we counted instead.
We analysed 999 reported US visa interviews — 2,753 individual questions and 1,197 document requests — spanning interviews from 2024 through August 2026, across H-1B, F-1, B-1/B-2, H-4, J-1 and K-1. Three findings contradict the standard advice.
1. The interview is far shorter than you have been told
Prep guides tell you to rehearse eight to fifteen questions. In the reports we counted, 81% of interviews involved five questions or fewer. The median was two. Only 4% ran past ten.
| Questions asked | Share of interviews | What it means |
|---|---|---|
| 1-2 | 50% | Half of all interviews are two questions or fewer |
| 3-5 | 31% | The typical substantive interview |
| 6-10 | 15% | Usually a consultancy case or a document discrepancy |
| 11+ | 4% | Rare — something specific is being worked through |
One caveat worth stating plainly: people write down the questions they remember, so these counts are a floor, not a transcript. But the shape holds — the long interrogation you are preparing for is the exception, not the norm. A long interview is a signal, not a default: in this data it correlates with third-party or vendor placements, a mismatch between an answer and the DS-160, or a prior refusal.
2. Social media went from a footnote to the defining question
This is the clearest before-and-after in the entire dataset. Comparing H-1B interviews reported before the universal social-media vetting rule took effect on December 15, 2025 against those after it:
| Topic | Before (n=103) | After (n=144) | Change |
|---|---|---|---|
| Social media handles | 2% | 67% | +65 pts |
| New visa or renewal | 3% | 15% | +12 pts |
| Threats or harm at home | 0% | 10% | +10 pts |
| Who you work for | 23% | 30% | +7 pts |
| Your job title | 21% | 14% | -7 pts |
| Where you will work | 25% | 14% | -11 pts |
Note what did not happen. Employer, salary and job duties are asked at roughly the same rate as before, and two topics actually fell. The median number of questions did not budge — three, before and after. The interview did not get harder or longer. It acquired one new topic.
That distinction matters for how you prepare, and it is why we trust this comparison. When a shift is caused by people simply writing longer posts, every category rises together. Here some rose and others fell, which is what a real change in questioning looks like.
The other thing to understand: social media is usually not a question you can answer wrongly. In most reports it is an instruction — confirm the handles on your DS-160, set the accounts to public. The failure mode is a typo in a DS-160 profile URL or a locked account, not a bad answer.
3. A question that did not exist a year ago
The strangest finding. Some version of these two questions appears 39 times in our data, and every single instance is after December 15, 2025:
“Do you have any threats in India?”
“Do you feel threatened about returning?”
Zero instances before the rule. What makes it notable is that it ignores visa category entirely — it shows up in H-1B, B-1/B-2, F-1, J-1 and H-4 interviews alike. One applicant noted that it was being asked at nearly every window that day regardless of what people were applying for.
It is a screening question, and the honest answer is the right one. It is worth knowing it is coming purely so it does not catch you off guard.
4. Officers ask for documents more often than they ask hard questions
Across all 999 interviews, document requests were the second most common event after social media. For H-1B specifically, the two documents most often asked for at the window were the I-797 approval notice and the LCA — carry both even when the appointment letter does not list them.
What actually happens after an H-1B interview now
Of 144 post-rule H-1B interviews in our data, 110 ended in a 221(g) and 27 were approved outright. A 221(g) is now the default outcome, not a red flag — and in tracked cases the majority resolve within days. We break the timelines down post by post in our odds by consulate analysis, and decode the CEAC status flips in the CEAC decoder.
See the questions for your visa and your consulate
The full ranked breakdown is live: interview questions by visa type and consulate, with the exact wording applicants remembered hearing. H-1B applicants can go straight to the H-1B list, or to the Hyderabad and Chennai breakdowns where the samples are large enough to differ meaningfully from each other.
Self-reported community data, not legal advice. These are applicants' recollections aggregated and stripped of identifying detail — a ranking of what is likely, not a guarantee of what you will be asked. Sample sizes are stated on every breakdown.