Cancer-related maternal mortality is a rare event. We report the first population-based study of this issue using data collected by the Committee on Maternal Welfare of the Massachusetts Medical Society between 1954 and 1985. The incidence of cancer-related maternal mortality during the study period fell from 3.16 to 0 per 100,000 live births. The most common cancer-associated maternal deaths were due to central nervous system tumors and hematological cancers. To determine the effects of pregnancy on cancer mortality, we compared our data with figures from the Connecticut Register of Mortality for Women aged 15-44. In the pregnant group there was a significantly higher incidence of mortality due to central nervous system tumors and a significantly lower incidence of mortality due to breast cancer. The data suggest that pregnancy may not be contraindicated for a woman with a history of breast cancer, but may be contraindicated for a woman with a history of a central nervous system tumor.