A short food frequency questionnaire (FFQ) to assess fruit and vegetable (F&V) intake was validated. Forty-nine adults and fifty-one adolescents (12-18 years old) completed the FFQ at home, and subsequently kept diet records for seven successive days. Mean daily intake of F&V was overestimated by the FFQ as compared to the 7-day diet records. For adults, spearman correlations of at least 0.5 were observed between the two methods for intake of total fruit and intake of citrus fruit. For adolescents, acceptable spearman correlations (0.53-0.64) were observed between the two methods for total F&V intake, total fruit intake and consumption of fruit juice. Low correlations (0.22-0.35) between the FFQ and the diet records were found for vegetable intake. Relatively large percentages (22-37%) of respondents were incorrectly classified by the short FFQ as eating according to the Dutch recommendations for fruit and vegetable intake. Based on the results of the present study as well as an overview of the results of 34 earlier validation studies of F&V FFQ's, it is concluded that the available FFQ's have only limited capability to make valid assessment of F&V intake levels. In spite of the practical advantages of the short FFQ tested in the present study and the similar validity scores as compared with other often used F&V FFQ's, it needs further revision, especially to be used for valid measurement of vegetable intake as well as total F&V intake among adults.