Abstract Objective To develop a ‘fruit self-efficacy’ (FSE) instrument and a ‘vegetable self-efficacy’ (VSE) instrument for secondary-school students in the Netherlands and to test the reliability and validity of these instruments. Design Methodological research. Setting Seven secondary schools in the Netherlands. Subjects Students (11–19 years of age) completed the two instruments in the classroom ( n 466). Out of the original subject group, 106 students completed the instruments a second time. Results The relevance of the twenty-two items of the two developed instruments was evaluated on two occasions by four experts. This procedure resulted in seventeen items for both instruments. Further analysis was carried out on the basis of the secondary-school students’ answers. Factor analysis identified two unidimensional instruments. Cronbach's α was 0·94 for the FSE instrument and 0·95 for the VSE instrument. The intra-class correlation coefficient between the test and the retest for both instruments varied between 0·33 and 0·84 ( P < 0·05), depending on how the tests were taken (during class or on the student's own initiative), and the order in which the tests were completed (started or finished with a different instrument used for similar research). The correlation between FSE and fruit consumption was 0·41 ( P < 0·01) and between VSE and vegetable consumption was 0·32 ( P < 0·01). Conclusions The two instruments are sufficiently reliable and valid to assess the FSE and VSE of secondary-school students in the Netherlands. Further research, in which the predominant ethnic minorities in the Netherlands are sufficiently represented, is required to confirm this conclusion.