Interviews were conducted with 303 men to determine aspects of work history and working environment that are related to low-back pain. Respondents included no pain, moderate pain, and severe pain samples. Data were obtained for 150 variables including task frequencies, weights lifted, lifting postures, type of floor surface, type of chair support, pushing, pulling, and carrying activities, driving vehicles, and attributions of pain occurrence to particular situations. Univariate analyses showed difference among the three pain groups for lifting method, having had lifting instruction, stretching and reaching activities, amount of sitting and twisting in seat, pushing materials of varied weights, and contributions of pain onset. Multivarlate regression analyses selected activities from (1) standing and lifting, (2) carrying, pushing, and pulling, (3) sitting, (4) driving vehicles, and (5) using vibrating equipment that jointly explained pain group variance. The significant variables from each domain were used in second-level multiple regressions to explain differences between no pain and pain, and between moderate and severe pain. Computed weights for pulling, lifting method, weight carried, reaching and stretching, chair support, number of times/day in-out of chair, and heavy vibration exposure explained 27% of variance between moderate and severe pain groups. Weights for automobile driving, number of times/day in-out of vehicle, number of lifts/day, maximum weight pulled, sitting with feet on floor, and exposure to heavy vibration explained 23% of the variance between no pain and pain groups