The relation between serum cholesterol concentration and mortality was studied prospectively over 11 years in 630 New Zealand Maoris aged 25-74. Serum cholesterol concentration was measured at initial examination in 1962-3 in 94% of the subjects and whether each was dead or alive was determined in 1974. The causes of death were divided into three categories: cancer, cardiovascular disease, and "other." The Mantel-Haenszel method of analysis of survivorship data showed a significant inverse relation between serum cholesterol concentration and overall mortality in men (x 2/2 = 11.6; p = 0.003) and women (x 2/2 = 7.6; p = 0.02) with odds ratios of 2.3 and 1.9 respectively. Similar significant inverse relations were found for cancer and "other" causes of death. These relations remained significant when baseline age, systolic blood pressure, and the Quetelt index were controlled in Cox's proportional hazards regression model. The results of this study provide evidence for a potentially deleterious effect of low serum cholesterol concentration. Hence, further research is needed before indiscriminate efforts are made to lower serum cholesterol concentrations in New Zealand Maoris.