The relationship of unemployment experienced between the ages of 16 and 33 years with smoking, alcohol consumption and obesity was examined in 2, 887 men who were members of the 1958 longitudinal British birth cohort study (NCDS). Cigarette smoking, alcohol consumption, measured as units consumed in the past week and as problem drinking using the CAGE questionnaire and the body mass index (BMI) were measured at age 33 years. Both the amount of unemployment accumulated between the ages of 16 and 33 years and recent unemployment experienced in the year prior to interview at age 33 years were examined. When compared with men who had never been unemployed, the adjusted relative odds amongst men with over three years of accumulated unemployment (after adjustment for possible confounding socioeconomic and behavioural factors measured prior to unemployment) were 2.11 (95% Cl: 1.42–3.12) for smoking, 2.13 (95% Cl: 1.32–3.42) for a low BMI and non-significant for a high BMI; 1.52 (95% Cl: 1.04–2.24) for no alcohol consumed; non-significant for high alcohol consumption, but 2.15 (95% Cl: 1.39–3.33) for problem drinking. Men who had experienced unemployment in the year prior to the interview, compared to those who had not, after adjustment, were significantly more likely to smoke (RO 2.92, 95% Cl: 2.13–4.01), drink heavily (RO 1.73, 95% Cl: 1.18–2.54) and to have a drink problem (RO 2.90, 95% CI: 1.99–4.21). Unemployment may play a significant part in establishing life-long patterns of hazardous behaviour in young men.