This study assessed the relationships of physical job demands (PJD), smoking, and alcohol abuse, with premature mortality before age 70 (PM-70) among the working or inactive population. The sample included 4,268 subjects aged 15 or more randomly selected in north-eastern France. They completed a mailed questionnaire (birth date, sex, weight, height, job, PJD, smoking habit, alcohol abuse (Deta questionnaire)) in 1996 and were followed for mortality until 2004 (9 yr). PJD score was defined by the cumulative number of the following high job demands at work: hammer, vibrating platform, pneumatic tools, other vibrating hand tools, screwdriver, handling objects, awkward posture, tasks at heights, machine tools, pace, working on a production line, standing about and walking. The data were analyzed using the Poisson regression model. Those with PM-70 were 126 (3.81 per 1,000 person-years). The leading causes of death were cancers (46.4% in men, 57.1% in women), cardiovascular diseases (20.2% and 11.9%), suicide (9.5% and 7.1%), respiratory diseases (6.0% and 4.8%), and digestive diseases (2.4% and 4.8%). PJD3, smoker, and alcohol abuse had adjusted risk ratios of 1.71 (95% CI 1.02-2.88), 1.76 (1.08-2.88), and 2.07 (1.31-3.26) respectively for all-cause mortality. Manual workers had a risk ratio of 1.84 (1.00-3.37) compared to the higher socio-economic classes. The men had a two-fold higher mortality rate than the women; this difference became non-significant when controlling for job, PJD, smoker and alcohol abuse. For cancer mortality the factors PJD3, smoker, and alcohol abuse had adjusted risk ratios of 2.00 (1.00-3.99), 2.34 (1.19-4.63), and 2.22 (1.17-4.20), respectively. Health promotion efforts should be directed at structural measures of task redesign and they should also concern lifestyle.