Oral Proficiency Interview Sample Questions

How the OPI Actually Works in Practice

Most people think you just pull up a list of questions and start grading someone's English. It doesn't work that way. The Oral Proficiency Interview Sample Questions you find online are useful reference material, but they're not a script you read from. The ACTFL guidelines don't require you to ask every candidate the same questions. The interviewer adapts based on what the person says. That's the whole point.

The interview typically runs 30 to 45 minutes. You start easy. "Tell me about your family" or "What did you do last weekend?" These are warmup questions that establish a baseline. From there, you move up in complexity based on the candidate's responses. If they handle a simple question with lots of detail and correct grammar, you push them to a more demanding prompt. If they struggle, you ease back down. The oral proficiency interview sample questions are really just prompts designed to elicit specific language functions at specific proficiency levels. Here's how the progression actually looks when you're running one of these interviews. The questions below map to ACTFL descriptors, which is what most institutions use. Novice High / Intermediate Low:

"Describe your daily routine." "What are your favorite foods and why?" "Tell me about a recent trip you took." Candidates at this level are working with sentence-level language. They can handle basic questions but will stall when asked for opinions or abstract topics. Intermediate Mid to High: "What are the advantages and disadvantages of studying abroad?" "Describe a problem you had and how you solved it." "What would you do if you won a lot of money?" These questions require paragraph-length responses. The candidate needs to sustain a narrative or give a supported opinion. This is where most people plateau.

Advanced Low to Mid: "Some people believe that social media has done more harm than good. What is your position?" "Discuss the impact of automation on the job market." "If you could change one thing about the education system, what would it be and why?" At this level you're looking for sustained discourse, effective negotiation of meaning, and minimal hesitation from language processing rather than thinking about what to say next. Advanced High and above:

These candidates don't really need structured questions. You have a conversation. You throw out a topic and see if they can handle abstraction, hypothesis, and nuance. "How might the definition of success differ across cultures?" "What responsibilities do individuals have versus governments when addressing climate change?" You're listening for fluent, precise language used in a natural, unstructured way.

What Nobody Tells You About Running These Interviews

The biggest problem I ran into early on was over-relying on the question list. I'd go through them like a checklist, and the interview felt robotic. The candidate would recover their composure after the first few prompts and then crush everything afterward. But I wasn't getting a true picture of their ability because I hadn't pushed them hard enough before they found their rhythm. I spent maybe four or five minutes establishing a baseline before I should have been pushing. Now I spend more time on the opening, probing deeper into simple answers. If someone says "I went to the store," I follow up with "What did you buy? Why did you choose that store? How does that store compare to others in your area?" That gets me to their real level faster. Another issue that caught me off guard: accented speech. A candidate with a strong accent can still be Advanced Low. The OPI isn't a pronunciation test. I've seen people mistakenly rate someone down because they were strain to understand the accent, not because the language was deficient. Make a note early on whether you're struggling with comprehension due to accent versus grammatical or lexical breakdown. Those are different problems. The trickiest edge case I encountered involved a candidate who was performing well on familiar topics but completely shut down when I shifted to a hypothetical. She was handling intermediate-high prompts fine, giving clear narratives and opinions. Then I asked her something like "What would you do if you were offered a job in another country?" Her language fractured. She kept circling back to simpler structures. I initially thought she was slipping toward intermediate mid, but I recognized the pattern. She had strong interactive competence but weak presentational/discursive ability. That's a real thing, and it comes up more often than you'd expect. The workaround was to give her scaffolding. Instead of a broad hypothetical, I broke it down: "First, imagine you received the offer. What would your immediate reaction be? Then, who would you talk to? What factors would you consider?" That approach revealed her actual ceiling, which ended up being solid advanced low. Without the scaffolding, I would have rated her significantly lower than she warranted.

Common Mistakes That Tank Your Ratings

Raters tend to be too generous with fluency. A candidate who speaks slowly but accurately should not rate higher than someone who speaks fluidly with frequent grammatical errors. Slow speech isn't the same as disfluent speech. Disfluency means language breakdowns, self-correction, restarting sentences. If they're just speaking deliberately, that's a different thing entirely. Another trap is confusing vocabulary range with proficiency. Someone who uses big words incorrectly is not advanced. They're precisely demonstrating they don't have control at that level. I once had a candidate who dropped words like "ephemeral" and "juxtaposition" into responses where they didn't belong. It sounded impressive on a first listen but broke down under scrutiny. The language wasn't serving clear communication. The scoring rubric itself has some fuzzy areas. The line between Intermediate High and Advanced Low is genuinely rough. Both levels can handle connected sentences and paragraphs. The difference is consistency and the range of contexts the candidate can navigate. A rating of half a level below where the person actually sits is very common and usually doesn't matter much for placement purposes. Being accurate to within a full level is the realistic target, not perfect precision.

Also worth noting: the OPI has real limitations. It's a speaking and listening assessment only. Someone who's Advanced Low in conversation might be completely unable to write a coherent paragraph. The OPI won't tell you that. It also tends to favor candidates who are comfortable with impromptu speech. People with test anxiety or cultural backgrounds where direct conversation with strangers isn't the norm often underperform relative to their actual language ability. If you're using OPI scores for high-stakes decisions like admissions or hiring, pair it with a writing sample. Don't rely on a single data point. There are also alternatives worth considering. The STAMP test gives you a quick computer-adaptive option that covers all four skills. ACTFL's OPIc (computer-delivered) lets you do the interview remotely, though some raters feel it's slightly harder to probe and redirect in that format. For most purposes the face-to-face or phone OPI is still the gold standard, but the OPIc has gotten significantly better over the past few years and it's far more scalable. If you're preparing to conduct these interviews yourself, the best resource is still the ACTFL Proficiency Guidelines document. It's dry reading, but it's the actual framework every rater is working from. The sample questions are secondary. Understanding what each level descriptor means in practice matters more than memorizing prompts.

What to Download and Use

ACTFL provides sample questions and rater training materials on their website. The official OPI rater certification requires going through their training modules, which walk you through recorded interviews and practice rating them. It's not optional if you want to be a certified rater. The free materials they publish are decent starting points but they won't fully prepare you. I'd recommend going through at least two practice cycles with a mentor or peer before you cut loose on real candidates. Rating calibration drifts fast when you're doing these solo.