A repeated measures design, with randomly assigned intervention and control groups and multiple sources of information on each participant, was used to examine whether changing the method of delivery of a school's homework program in order to better meet the students' needs for autonomy, relatedness and competence would lead to more positive student attitudes to homework, and whether there would also be a positive change in overall motivation. The participants were 104 male students aged 10 to 12 years who attended a single sex high school. There was no overall intervention effect on motivation; however, the intervention appeared to have a protective effect on the quality of motivation.