Abstract Background To unravel the effect of school-based nutrition education, insight into the implementation process is needed. In this study, process indicators of Taste Lessons (a nutrition education programme for Dutch elementary schools) and their association with changes in behavioural determinants relevant to healthy eating behaviour are studied. Methods The study sample consisted of 392 Dutch primary school children from 12 schools. Data were collected using teacher and child questionnaires at baseline, and at one and six months after the intervention. Multilevel regression analyses were conducted to study the association between dose, appreciation and children’s engagement in interpersonal communication (talking about Taste Lessons with others after the lessons), and change in knowledge, awareness, skills, attitude, emotion, subjective norm and intention towards two target behaviours. Results With an average implementation of a third of the programme activities, dose positively predicted change in children’s subjective norm of the teacher after one month. Teachers and children highly appreciated Taste Lessons. Whereas teacher appreciation was inversely associated, child appreciation was positively associated with children’s change in awareness, emotion and subjective norm of teachers after one month and in attitude and subjective norm of parents after six months. Interpersonal communication was positively associated with children’s change in five determinants after one month and in attitude and intention after six months. Conclusions The implementation process is related to the programme outcomes of Taste Lessons. Process data provide valuable insights into factors that contribute to the effect of interventions in real-life settings.