In a rural area of Senegal, 1500 cases of measles and 98 deaths (6.5%) were registered during a 4-year period (1983-1986). For children less than 5 years old, the acute case fatality ratio (CFR) was 9.6%. Compared with index cases who contracted measles in the village, secondary cases infected within the compound had a significantly higher mortality (odds ratio [OR], 2.9; 95% confidence interval [CI], 1.3-6.1). The CFR of secondary cases increased with the closeness of contact with the index cases; the ORs were 1.9 (CI, 0.6-6.0) for children living in the same compound but not the same household, 2.3 (CI, 1.0-5.7) for children living in the same household but not in the same hut, and 3.8 (CI, 1.7-8.4) for children living in the same hut. In large compounds with many cases, the CFR increased exponentially through the succession of generations; the ORs were 2.3 (CI, 1.0-5.2) in the second generation, 3.7 (CI, 1.1-3.0) in the third generation, 5.5 (CI, 1.7-18.1) in the fourth generation, and 16.1 (CI, 5.6-46.3) in the fifth or later generations. Postmeasles mortality through the year following measles infection was also significantly related to the intensity of exposure. Differences in exposure may be a major determinant of child survival, both at time of acute disease and for the long term impact of measles infection.