To better describe patterns of nonoccupational morbidity among construction laborers and their dependents, two health insurance plans organized by local unions of the Laborers' International Union of North America provided their medical claims data for 1989. The observed numbers of hospital admissions were compared with the numbers expected, based on the age-sex-specific hospital discharge rates from the 1989 National Hospital Discharge Survey. Standardized morbidity ratios thus obtained showed excesses for alcohol and drug dependence, complications related to pregnancy, and several other conditions. Medical claims data are a very useful resource in epidemiologic and medical care research, but their use poses numerous challenges, mainly related to the accuracy of diagnostic recording, problems in comparing different health insurance plans, and confounding factors due to health insurance largely being a condition of employment. Nevertheless, the use of these data can provide specific hypotheses for further study.