This study assessed the relative validity of the following non-laboratory measures of compliance with a fat-controlled diet by hypercholesterolemic males: quantitative and qualitative measures of compliance derived from three-day food records and structured interviews, and summary ratings by nutritionists and interviewers. The quantitative measures of consumption referred to: saturated fat, linoleic acid, polyunsaturated fat, P/S ratio and cholesterol. The non-laboratory measures were assessed using the following two laboratory measures as criteria of validity: serum cholesterol concentration and the proportion of linoleic acid among the fatty acids of the serum cholesteryl esters. The results showed that, with respect to quantitative measures, the food record and interview manifested approximately equal validity, with the former marginally more valid. However, under certain conditions qualitative measures, such as patients' self-ratings, were about as valid as the more rigorous and expensive quantitative measures. Characteristics of the patient influenced the validity of the various measures. Validity was enhanced by the simultaneous use of certain combinations of measures.