Free 20-question cheating test with an instant score out of 20 and a plain-language read on what it means. No signup, nothing stored, works in your browser.

Twenty yes/no questions across five categories. Every "yes" is one point. Answer honestly — the result is worthless if you shade your answers toward what you're hoping to hear, in either direction.
Most cheating quizzes online just add up warning signs. That's the wrong shape for this problem. A score of 10 concentrated entirely in the phone category is far weaker evidence than a score of 8 spread across phone, time, appearance and money — because a single category has a single explanation, and four categories changing at once usually don't.
Someone who suddenly guards their phone might have started a new job with a strict NDA. Someone who suddenly guards their phone and has unexplained gaps in the evening and bought new clothes they don't wear around you and switched to paying cash has a coordination problem, and coordination is what a second relationship requires.
That's why the five categories in this test are deliberately unrelated to each other. The signal you're looking for isn't intensity in one area — it's simultaneity.
It cannot tell you that your partner is cheating. No behavioral checklist can, and any test claiming otherwise is lying to you for engagement. What this test does is convert a vague feeling into a number you can reason about — which is genuinely useful, because vague feelings are impossible to act on and easy to dismiss.
A high score has at least four other common explanations, and they're worth taking seriously before you conclude anything:
Each of those deserves an entirely different response than infidelity does. That's the practical argument for getting one hard fact before you decide which story you're in.
This is the step almost everyone skips and almost everyone regrets. An accusation backed by a feeling gives the other person the easiest possible exit: they get to make the conversation about your suspicion rather than their behavior. If they are hiding something, you've also just told them exactly what to hide better.
Not surveillance — one fact. The most checkable thing in 2026 is whether their face is attached to an active dating profile, because that's the one piece of the picture that lives on a public platform rather than inside their phone. Run an anonymous photo search →
Calm room, no children present, no deadline. Lead with what you've observed rather than what you've concluded. If you have a fact, state it once and then stop talking — the silence does more work than the accusation would.
Leaving, therapy, and a structured pause are all legitimate outcomes. What isn't legitimate is another six months of the state you're in right now. Prolonged uncertainty does more measurable damage to people than a hard answer does — the research on ambiguous loss is fairly consistent on that point.
The twenty items come from the behavioral patterns that recur most consistently in three places: documented infidelity cases, the relationship-psychology literature on secrecy and disclosure, and the sequence people describe in retrospect once they know what happened. We weighted nothing — every item is one point — precisely because weighting invites the author's assumptions in, and the honest version of this tool is a count plus an interpretation, not a hidden model pretending to be a diagnosis.
The bands are calibrated conservatively. A score of 5 does not mean "probably cheating," and we don't tell you it does. It means you're in the range where watching more rarely resolves anything and one objective check usually does.
This test is free, requires no account, and stores nothing you answer — the scoring happens entirely in your browser. If you run a relationship blog, a therapy practice site or a forum where this question comes up constantly, the embed snippet above is there for you to use.
It accurately measures one thing: how many of the 20 behavioral patterns associated with infidelity you have actually observed. It cannot measure whether your partner is cheating, because behavior alone never proves that. Treat the score as a way to see whether what you've noticed forms a pattern across several unrelated categories, which is far more meaningful than any single warning sign on its own.
Twenty yes/no questions, split across five categories: phone and device behavior (5), emotional distance (4), appearance and routine (3), time and whereabouts (4), and money and instinct (4). Each yes is one point, so scores run from 0 to 20. It takes about three minutes.
No score means that. A score of 0–4 shows very few indicators, 5–9 is a moderate cluster, 10–14 is strong, and 15–20 means the pattern is present across every category. Even at 20, depression, burnout, a hidden financial problem or an undisclosed health issue can produce almost the same profile — which is exactly why the test points you toward one objective check rather than a conclusion.
Yes. There is no signup, no email required, and nothing you answer is sent anywhere — the scoring runs entirely in your browser. If you share your result, the link carries the score number only, never your answers.
Yes. There is a copy-paste embed snippet on the page, free to use, which renders a small card linking to the test. It includes an attribution link back to PartnerCheck. No permission or API key needed.
If you scored in the moderate range or above, the useful next step is one objective fact rather than more observation — most people get stuck cycling between suspicion and self-doubt for months. The most checkable fact in 2026 is whether your partner's face is attached to an active dating profile, since that lives on a public platform rather than inside their phone. Then plan the conversation, then decide deliberately.