Hospitalization rates for chronic medical conditions vary across small areas and are associated inversely with community income. The authors studied whether variation in hospitalization rates can be attributed to differences in physician practice style. Using census and hospital discharge data, hospitalization rates were calculated for asthma, congestive heart failure, and diabetes in 40 medical service areas in California. The authors surveyed a random sample of 1,530 emergency physicians, general internists, and family and general practitioners in these areas, and measured clinical admission threshold by asking physicians whether they would hospitalize patients presented in 15 vignettes of graded severity. The authors measured social admission predisposition by asking how physicians' admission decisions would be influenced by social characteristics that increase patients' vulnerability to illness, including homelessness and drug use; 1,090 physicians responded (71%). There was significant variation across areas in both the clinical (P < 0.0001) and social (P < 0.001) admission scores. Variation in hospitalization rates correlated with physicians' clinical (r = .34, P = 0.03) and social (r = .36, P = 0.02) admission scores. However, in a multiple linear regression analysis that included community sociodemographic factors, physician practice style was not associated significantly with hospitalization rates. Physician practice style varies across areas, but does not explain variation in admission rates for chronic medical conditions after adjusting for community sociodemographic factors. Using methods such as practice guidelines or utilization review to re-set physicians' threshold for admission may not be effective in reducing hospitalizations for chronic medical conditions.