Substantial dietary underreporting questions the validity of dietary measurements of energy intake (El). The present study compares the value of a 7-day prospective dietary record (7dDR) with a computer program-based diet history (DH). 7dDR and DH were performed in 50 non-obese subjects (33 females, 17 males, mean age 26.1 years, BMI 18.9-29.6 kg/ m2) using total energy expenditure (EE = sum of resting metabolic rate as measured by indirect calorimetry plus energy expenditure derived from an activity protocol) as standard for the validity of data on EL El was 2,206 (728-3,646) kcal/day for 7dDR and 2,398 (566-4,764) kcal/day for DH. There was an association between El for 7dDR and El for DH (r = 0.6, p < 0.0001). Underreporting [i.e. a difference between El and EE (ΔE = EI-EE)] of 20% or more was seen in 48% (7dDR, mean -1,047 kcal/day, range -616 to -1,895, or -38.8% of EE) or 48% (DH, -1,151 kcal/day, -594 to -2,057 kcal/day, or -42.3% of EE). Considerable differences were found between ΔE for 7dDR and ΔE for DH (mean 603, range 26-2,033 kcal/ day), and only 34% of underreporting subjects were identified by both dietary measurements. It is concluded that at the individual level dietary underreporting is influenced by the dietary assessment to