The aim of the research was to evaluate the impact of intervention materials, designed to enhance self-efficacy and anticipated regret, on contraceptive behaviour and antecedents of contraceptive use in a sample of adolescents. It was hypothesised that materials designed to enhance self-efficacy and anticipated regret would lead to improvements in outcome measures compared with controls. A 4(intervention condition) × 3(time) mixed design was used to assess the impact of intervention materials. Participants (N = 414) were recruited from five secondary schools in the north of England. They were assigned to an active control group, an anticipated regret (AR) manipulation, a self-efficacy (SE) manipulation or both AR and SE manipulations. Outcome measures included psychological antecedents of contraceptive behaviour change, intentions and behaviour. Multivariate analysis of variance (MANOVA) revealed increases across several outcome measures over time (F[14,287] = 8.99, P < 0.001, ηp 2 = 0.305) including intentions, but these did not differ by condition (F[42,852] = 1.35, P = 0.07, ηp 2 = 0.062). There was evidence that the questionnaires may have caused reactivity in participants. Amongst sexually active participants with relatively low levels of intention to use contraception at the outset, increases in several outcome measures including intention and behaviour were observed (F[3,35] = 10.359, P < 0.001, ηp 2 = 0.47). Findings support the potential for effective delivery of behaviour change theory-driven interventions in classroom settings. The possibility that the questionnaires may have acted as a form of intervention contributes to recent discussion of this issue in the literature, and the findings also strengthen the case for post-decisional and behavioural skills interventions to enhance behaviour amongst those already motivated to use contraception.