Competition often is viewed as a mechanism for controlling cost. Competition may work well in urban areas with many providers; competition may not exist in rural areas with few providers. The authors use the empirical framework developed by Bresnahan and Reiss to analyze the entry behavior of physicians into local markets to determine the level of physician supply consistent with competitive behavior. The study estimates entry patterns for total and specialty physicians located in nonmetropolitan health service areas using longitudinal data. The authors find a surprising drop in the population increments necessary for entry by the second provider, possibly due to the unattractiveness of being the solo physician in an area. Subsequent population increments stabilize at three to five physicians. Since more than 93 percent of the U.S. population lives in areas that can support three to five physicians, competition between physicians through mechanisms such as managed care may be feasible.