The aim of this paper is to evaluate the use of the Mainz Dortmund dose model in a case-control study of lumbar spine disease. In two practices and four clinics we recruited 229 men suffering from symptomatic osteochondrosis or spondylosis of the lumbar spine which had been confirmed radiographically. A total of 197 controls were recruited (107 population controls; 90 hospital-based controls with urolithiasis in whom osteochondrosis/spondylosis of the lumbar spine had been radiographically excluded). Data were collected in a structured personal interview and analysed using logistic regression with regard to age, region, nationality and diseases affecting the lumbar spine. The Mainz Dortmund dose model was used to calculate the cumulative strain on the lumbar spine. In the analysis of patients with symptomatic osteochondrosis/spondylosis and controls we found a dose-response relationship between the cumulative strain on the lumbar spine from carrying/lifting and extreme forward bending. The analysis of occupational groups and that based solely on the individual control subjects' exposure assessments yielded high risk estimates. These results suggest that differential recall was not an adequate explanation for the relationship between physical workload and symptomatic osteochondrosis/spondylosis. Even when - unlike in the original Mainz Dortmund close model - no minimum exposure limits for inclusion in the calculation of the cumulative strain on the lumbar spine were defined, we found large odds ratios for expoures far below the lower risk limit of 25 × 106 Newton-hours. To constitute a useful tool in risk assessment, we recommend all carrying/lifting activities with a minimum weight of 5 kg be included in the calculation of cumulative strain on the lumbar spine, and the lower risk limit of the original Mainz Dortmund dose model be neglected. Larger epidemiological studies need to be carried out to determine the relationship between carrying/lifting or extreme forward bending and lumbar spine disease.