The most recent evidence from population-based cohort studies confirms the association between malignancy and dermatomyositis and polymyositis. These studies show an even stronger association between polymyositis and malignancy than previous studies, suggesting less misclassification. This is particularly true of one study that used pathologic criteria to distinguish between myositis subtypes. Recent data also confirm that the association for dermatomyositis and polymyositis is not purely caused by diagnostic suspicion or surveillance bias. More data are still required to determine individual cancer risks, although it appears that ovarian and lung cancer are associated with dermatomyositis while lung cancer and non-Hodgkin's lymphoma are associated with polymyositis. An association between malignant disease and inclusion body myositis has also been verified for the first time. Of interest, too, is the increasing number of reports documenting cases in which the clinical course of the myositis mirrors that of the cancer, supporting the notion that in some instances, myositis is a paraneoplastic disorder.