Aims. To evaluate the effects of a brief intervention to reduce excessive drinking, smoking and stress among police. Design. (1) Controlled intervention trial with pre and post-intervention assessment approximately 8 months apart; (2) focus group identification of relevant factors). Setting, participants. Assessment was carried out of 954 NSW (Australia) police at 19 stations within two matched districts in the Sydney metropolitan area. Five focus groups were carried out with 43 randomly selected police from the matched districts. Measurements. Weekly alcohol consumption and binge drinking, smoking and symptoms of stress were measured by a self-administered Health and Fitness Questionnaire. Recorded responses to set questions provided qualitative data. Results. Participation was high (89%) at both quantitative assessments. Alcohol consumption, particularly among men, was high at both baseline and follow-up assessments, although comparisons between groups across occasions showed no significant intervention effects. Excessive drinkers and those reporting moderate to severe stress levels reported more sick leave days ( p < 0.05, p < 0.05). A significant increase in awareness of alcohol policies in the work-place showed in both experimental and control groups over time (p < 0.01). The percentage of smokers declined significantly in both intervention and control groups. Overall, women had significantly more symptoms of stress than men. Only 20% of police thought they would seek advice from work-place staff about alcohol consumption, 14% for smoking and 61% for stress. In the qualitative study, employees generally distrusted their organization's involvement in health unless work performance was affected. Seeking professional assistance for life-style issues was viewed as a sign of weakness. Alcohol use was seen as a way of obtaining information or group membership, self-medication and socializing. Conclusions. The brief interventions did not produce significant improvements in three life-style factors beyond positive trends in alcohol consumption among women and general reductions in smoking among both study groups. Combining quantitative and qualitative approaches helped identify interactive individual and organizational factors which influence behavioural and cultural norms.