Are science YouTube videos accurate?

Science is the one corner of YouTube that comes with a built-in expectation of accuracy — and for once, the data mostly rewards it. When we ran popular science and education explainers through the same automated credibility check that powers the Credibility Leaderboard, the category tied for the highest average substance score of any topic, at 86/100. It also produced four of the seven videos in the entire set that cleared 90, including the single highest score we recorded (95).

That is not an accident of the subject matter; it is a reflection of what good science communicators actually do. The top-scoring explainers earned their numbers by being specific and by showing their working — naming mechanisms, citing evidence, and demonstrating a result rather than narrating a conclusion. Our scoring rewards exactly that: specificity, sourcing, balance, accuracy, and depth. A video that says “studies show” scores lower than one that shows the study.

The caveat matters just as much as the good news. A high average does not mean every science video is safe to trust, and popularity is still not the signal — plenty of slick, high-view science content oversimplifies to the point of being misleading. The reason the best ones score well is method, not fame. When a science video skips the sourcing and leans on confident narration over demonstration, the score drops, popular channel or not.

So the practical takeaway is the same one that works everywhere else, just with a higher base rate of success: count how often the creator shows rather than tells. In science, the videos that demonstrate and cite are the ones that hold up — and they are the ones that sit at the top of the board.