A case-control study was conducted to examine possible relationships between human acute lymphoid leukemia and exposure to dairy cattle and drinking of raw milk. Two hundred twenty-three persons with acute lymphoid leukemia, diagnosed during the years 1969 to 1971 and 1973 to 1980 from the 87 most rural Iowa counties, were accessed from case records at the Iowa State Health Registry for participation in the present study. Each person and 2 matched controls were interviewed for history of residence, exposure to dairy cattle, and consumption of nonpasteurized dairy products. Two types of comparisons between affected persons and controls were done: the prevalence of bovine leukemia virus infection (as measured by serologic study) in dairy herds with which the affected persons and controls had either occupational contact or from which they had consumed raw milk and the density of dairy cattle in the townships where affected persons and controls lived. The bovine leukemia virus infection prevalence in dairy herds with which affected persons had contact was 20%, whereas the infection prevalence in the herds with which the controls had contact was 38%. The density of dairy cows in townships where affected persons resided was generally less than that in townships where controls resided. However, there was one exception; the density of dairy cows at 20 years before diagnosis was higher (589) in townships where affected adult female persons resided, compared with that in townships where controls resided (567).