The objective of the present study was to validate energy intake data, obtained by dietary history, in twelve elderly women aged 69-82 years. Energy and protein intakes were obtained using the dietary history method with a reference period of 30 d. Reported energy intake was compared with total energy expenditure (TEE) measured on two consecutive days in a respiration chamber. Reported protein intake was compared with mean N excretion from four 24 h urine collections. Mean reported energy intake was 7.2 (SD 1.5) MJ/d which was lower than TEE (P = 0.059). Reported protein intake was 64 (SD 13) g/d and lower than estimated protein intake (P = 0.053). The percentage underestimation was not related to body weight or percentage body fat. Subjects with a relatively high TEE or a relatively high estimated protein intake underestimated their energy intake to a greater extent. The discrepancy between reported energy intake and TEE was positively associated with the discrepancy between reported and estimated protein intakes. The results of this present study show an underestimation of energy intake of about 12% when using the dietary history method. Physical activity diaries completed in the chamber and during 4 d at home, as well as pedometer counts, indicated a higher level of physical activity in the free-living situation compared with the chamber situation. This suggests that the actual underestimation of energy intake may be even higher in this group of elderly women. These results have implications for the use of the dietary history method in, for example, epidemiological studies carried out in elderly subjects.