Examination of the case-records of women presenting to the Melanoma Unit at Sydney Hospital over the period 1961-71 has shown that women with pregnancies before the development of melanoma had a better survival-rate from melanoma than women without previous pregnancies. The known presence of fetal antigens on melanoma cells and immunisation against fetal antigens during pregnancy suggest an immunological explanation for these results. Exposure to fetal antigens during pregnancy may protect against the dissemination of melanoma cells bearing similar fetal antigens and thus increase the survival-rate. The incidence of melanoma in males and females was approximately equal, which suggests that immune responses to tumour-associated antigens may be more effective in preventing spread of tumours than in preventing their occurrence.