In this Draft, the authors describe the demographic, health and functional status factors associated with driving after age 50. Using data from the 1990 Panel Study of Income Dynamics, the authors compared characteristics of drivers and nondrivers over age 50. They then used logistic regression to predict the likelihood that persons who had previously driven would continue to drive and drive after dark. Results showed that when multivariate techniques were used to predict the likelihood of having stopped driving after age 50, most standard health status measures were not significant predictors. Instead, a combination of demographic variables (age, gender, education, and marital status), household composition, global self-perceived health, and three medical conditions (visual impairment, arthritis, and major neurological impairment) explained a large percentage of the variance in driving patterns. The significant predictors for driving after dark were similar. Those persons who did not drive largely relied on family and friends for their transportation needs. The authors conclude that sociodemographic factors and specific health conditions explain a substantial amount of the variation in driving patterns, and older drivers appear to self-regulate their driving patterns at least partly based on medical conditions.