After separating ambulatory visits into those made in connection with illness or injury and preventive visits, the utilization patterns of a sample of families and individuals are analyzed. Need, in terms of perceived health status and the numbers of acute and chronic conditions, price, and access are found to be the best predictors of visit rates, but their roles in illness and preventive visit rates are different. The methodologically relevant findings indicate that individual self-reports and independent individual observations are required to identify relationships hidden by family member data, such as that between hospital episodes and ambulatory visits. The substantive findings indicate a substitutive relationship between illness and preventive visits, lend further evidence for relatively low price elasticity for illness visits and show that membership in a closed panel health maintenance organization increases preventive visit rate while price has little or no effect on it. The tentative policy implication--that it is not so much price as the characteristics of the usual source of care which appear to determine preventive services utilization--is discussed in the context of potential biases inherent in the sample.