The objectives of this prospective study were to identify predictive factors for sick leave of 8 days or more due to low back pain (LBSL) and to compare them with predictive factors for low back pain with no or shorter sick leave (LB) in a cohort of French workers. The predictive factors for LBSL were a past history of low back pain (odds ratio [OR], 7.2; 95% confidence interval [CI], 4.1 to 13), a low employment grade (OR, 4.3; 95% CI, 1.7 to 11), heavy smoking (OR, 5.5; 95% CI, 2.3 to 13), a pain score different from zero (OR, 4.9; 95% CI, 2.5 to 9.7), required bending backward or forward at work every day repetitively (OR, 7.4; 95% CI, 2.3 to 23), overall social integration (OR, 2.0; 95% CI, 1.3 to 3.3), and low social support at work (OR, 3.4; 95% CI, 1.6 to 7.3). Low social support at work and bending backward or forward at work were more strongly associated with LBSL than with LB (P = 0.02 and P < 0.01, respectively). The implications of the results of this prospective study are that both the level of biomechanical exposure and the psychosocial work environment, especially social support, represents important dimensions to consider in the reduction of work absenteeism.