Questionnaire data from 211 adolescents and follow-up data recorded 18 months later were employed to test main effects and stress-buffering effects of negative life events, on-going stressors and social support from family and friends on mental health. Negative life events, change from baseline level of on-going adversities and social support all contributed significantly to subsequent symptom scores, although negative life events only reached borderline significance among boys. There was evidence in favour of the buffer hypothesis for boys: negative life events had a significantly stronger effect when social support from peers was low, and long-lasting adversities had a significantly stronger effect when social support from parents was low. Both these two-way interaction effects among boys were significantly different from the corresponding trends among girls. Since the scores on both the independent and dependent variables are based on subjective self-reports, the results may have been affected by various types of response bias. The probabilities of such bias effects are discussed.