Objective: To investigate the impact of school garden-enhanced nutrition education (NE) on children's fruit and vegetable consumption, vegetable preferences, fruit and vegetable knowledge and quality of school life. Design: Quasi-experimental 10-week intervention with nutrition education and garden (NE&G), NE only and control groups. Fruit and vegetable knowledge, vegetable preferences (willingness to taste and taste ratings), fruit and vegetable consumption (24 h recall 3 2) and quality of school life (QoSL) were measured at baseline and 4-month follow-up. Setting: Two primary schools in the Hunter Region, New South Wales, Australia. Subjects: A total of 127 students in Grades 5 and 6 (11-12 years old; 54 % boys). Results: Relative to controls, significant between-group differences were found for NE&G and NE students for overall willingness to taste vegetables (P , 0?001) and overall taste ratings of vegetables (P , 0?001). A treatment effect was found for the NE&G group for: ability to identify vegetables (P , 0?001); willingness to taste capsicum (P 5 0?04), broccoli (P 5 0?01), tomato (P , 0?001) and pea (P 5 0?02); and student preference to eat broccoli (P , 0?001) and pea (P , 0?001) as a snack. No group-by-time differences were found for vegetable intake (P 5 0?22), fruit intake (P 5 0?23) or QoSL (P 5 0?98). Conclusions: School gardens can impact positively on primary-school students' willingness to taste vegetables and their vegetable taste ratings, but given the complexity of dietary behaviour change, more comprehensive strategies are required to increase vegetable intake.