A patient presenting with interstitial lung disease was predicted to have underlying polymyositis when found to be anti-Jo-1 antibody positive. When myopathy developed later with elevated serum creatine phosphokinase and myoglobin levels, the electromyogram was nonmyopathic, and polymyositis was confirmed by muscle biopsy. The case illustrates the value of anti-Jo-1 as a 'marker' autoantibody for polymyositis, and the need for thorough investigation of patients with interstitial lung disease of uncertain etiology. This should include indirect immunofluorescence using a dividing cell substrate and tests for antibodies to extractable tissue antigens. The need to completely evaluate patients with dermatopolymyositis is also emphasised.