Common preemployment screening methods were evaluated for their ability to predict individuals at risk for occupational back problems. A total of 3020 workers from a Boeing Company factory in Washington State, participated in a questionnaire and physical examination. Screening involved questions on demographics, medical history with emphasis on pain problems, and psychosocial factors. Physical examination dealt with posture, reflexes, and musculoskeletal measurements. Of the volunteers, 1569 (54%) completed the questionnaires. During follow up, 279 reported back problems. Back injuries were reported by 136 (8.7%) of those who completed the questionnaire, and 13 (9.9%) of those who did not. The difference in injury rates was not significant. Data were analyzed by the product life table method and the Cox proportional hazards regression model. According to univariate analysis, younger age was a significant marker for predicting acute back pain. In men, more years of education was associated with higher risk, as was highest level of completed education. This association did not hold for women. Smokers were also at greater risk of injury. Significant past medical history predictors were doctor visits for any reason, pain medication use in the previous 2 years, hospitalization in the previous 10 years, history of back surgery, consultation with a chiropractor, and previous workers' compensation claim for back problems. Greater standing height in men, and increased weight in women were weak predictors. Asymmetric decreased reflex at the right ankle was significant in women, but not in men. Back symptoms on straight leg raising was significant for the general population. Multivariate analysis revealed that treatment for pain problems in the previous 2 years and seeing a chiropractor for any reason explained the predictive power of the other variables in men, as did younger age and greater number of hospitalizations over the previous 10 years in women. The authors conclude that if medical history regarding previous pain treatment is known, information from physical factors does not add any significant predictive value to the screening method.