This study reviews articles published in 2000 and 2001 about occupational physical activities as risk factors for osteoarthritis (OA) of the knee and hip. Six studies were identified in MEDLINE and EMBASE; five case-control studies and one cross-sectional study. Criteria for OA were: (on the waiting list for) joint replacement or (symptomatic) radiologic OA. Exposure was assessed by structured interviews or self-administered questionnaires. Job titles or details on specific physical activities were used as determinants. The authors determine that information bias and nondifferential misclassification, with respect to occupational risk factors, cannot be excluded. Relevant confounding factors were adequately taken into account. Even though response rates were high, nonresponse bias may still have an influence, although its impact cannot be assessed accurately. Within these constraints the studies gave further evidence to support the role of jobs and occupational physical activities on the occurrence of OA.