Surveys that ask people to recall what they ate over long periods, using infrequent questionnaires, do not provide accurate measurements of how much of specific foods they actually consumed over time.

We analyzed the available evidence on food frequency questionnaires and found that these surveys, which ask people to recall what they ate over weeks or months, do not accurately measure long-term intake of specific foods [1]. The evidence we’ve reviewed so far consistently suggests that memory-based reporting leads to significant gaps between what people say they ate and what they actually consumed. This is not because people are being dishonest—it’s because human memory fades, portion sizes are hard to estimate, and eating habits change day to day. Even when people try their best, the structure of these questionnaires makes precise tracking difficult. We saw no studies in our review that contradicted this finding. What we’ve found so far leans toward the idea that while food frequency questionnaires can give a rough idea of general eating patterns—like whether someone eats a lot of vegetables or很少 fruit—they are not reliable for measuring exact amounts of calories, nutrients, or specific foods over time. This matters because many health studies rely on these tools to link diet with outcomes like weight gain or disease risk. If the data is inaccurate, the connections drawn from them may be misleading. For someone trying to track their own eating habits, this means relying on a single questionnaire won’t give you a clear picture of your long-term diet. Better tools—like food logs kept daily or digital tracking apps—may offer more useful insights, though even those have limits.