OBJECTIVES: To test the hypothesis that work-related mechanical, psychosocial and physical environment factors would predict new-onset low back pain (LBP) in newly employed workers. METHODS: A total of 1186 newly employed workers were recruited from a variety of occupational settings. Those who were free from LBP at baseline were identified. Subjects were followed up at 12 and 24 months. Work-related mechanical, psychosocial and physical environment exposures were measured. Generalized estimating equations were used to assess predictors of new-onset LBP. RESULTS: New-onset LBP was reported by 119 (19%) and 81 (19%) subjects at 12 and 24 months, respectively. Several work-related mechanical exposures predicted new-onset LBP including lifting heavy weights with one or two hands, lifting heavy weights at or above shoulder level, pulling heavy weights, kneeling or squatting for 15 min or longer. Of the psychosocial factors examined, stressful and monotonous work significantly predicted symptom onset. In addition, hot working conditions and pain at other sites also predicted new-onset LBP. On multivariate analysis these risks were only moderately attenuated but the 95% confidence intervals excluded unity only for the latter, non-mechanical, exposures. CONCLUSION: In this cohort of newly employed workers, from a range of occupations, several aspects of the work-place environment, other than mechanical factors, were important in predicting new-onset LBP. These results emphasize that interventions aimed at reducing the occurrence of LBP are likely to be most successful if they intervene across these domains ; UI - 22752177DA - 20030718IS - 1462-0324LA - engPT - Journal ArticleSB - AIMSB - IM