The objectives of this study were (1) to develop a scale to measure patient preferences for using medical care, (2) to assess the reliability and validity of the scale, and (3) to examine factors predicting preferences. Preferences were defined along a continuum, anchored by self-treating preferences and care-seeking preferences. A nine-item scale was developed and mailed to a random sample of 3500 Wisconsin consumers age 50 and older. Ordinary least squares regression was used to examine whether preferences were predicted by demographic and health status variables. A 56.9% usable response rate was obtained. The Medical Care Preference Scale was unidimensional and had a Cronbach’s alpha of 0.879. Younger individuals, women, individuals in better health, and individuals from rural areas had significantly stronger self-treating preferences. Significant correlations between the preference scale and two measures of health care utilization provided evidence of predictive validity. Individuals with care-seeking preferences used an average of 1.98 more prescription drugs and had 0.50 more physician visits in the past month than individuals with self-treating preferences. The Medical Care Preference Scale should be a useful tool for research on health care utilization.