OBJECTIVE:To assess whether levels of dietary restraint are associated with mis-reporting measures of adiposity by middle-aged adults. DESIGN: Cross-sectional study of middle-aged men and women, the parents of a group of young adults followed up since birth. METHODS: In all, 631 couples were mailed questionnaires and asked to record their height, weight and waist circumference. A paper tape measure with instructions for use was attached. Couples also completed the dietary restraint section of the Dutch Eating Behaviour Questionnaire, and provided information on employment and lifestyle habits. A subsample of participants was then invited to attend a clinic where detailed anthropometric measures were taken. RESULTS: In all, 435 women (69%) and 332 men (55%) completed the questionnaire; of those invited, 182 (85%) women and 102 (61%) men attended a clinic session. Regression analyses showed that the dietary restraint score was associated with the mis-reporting of BMI by women (Po0.01), but not men (test for interaction with gender, P ¼ 0.11). In women, the difference between the measured and reported BMI increased by 0.36 kg/m 2 (0.11-0.61) per unit increase in restraint score. This association was independent of age, smoking, social class, slimming, exercise frequency or television viewing time, but was attenuated in models controlling for measured BMI. The dietary restraint score was not associated with mis-reporting of waist circumference in men or women. CONCLUSIONS: Dietary restraint score may be a useful tool for identifying individuals more likely to mis-report anthropometric measurements, although associations may vary by gender.