The Medicare prospective payment system (PPS) contains incentives for hospitals to improve efficiency by placing them at financial risk to earn a positive margin on services rendered to Medicare patients. Concerns about the financial viability of small rural hospitals led to the implementation of the Medicare Rural Hospital Flexibility Program (Flex Program) of 1997, which allows facilities designated as critical access hospitals (CAHs) to be paid on a reasonable cost basis for inpatient and outpatient services. This article compares the cost inefficiency of CAHs with that of nonconverting rural hospitals to contrast the performance of hospitals operating under the different payment systems. Stochastic frontier analysis (SFA) was used to estimate cost inefficiency. Analysis was performed on pooled time-series, cross-sectional data from thirty-four states for the period 1997-2004. Average estimated cost inefficiency was greater in CAHs (15.9 percent) than in nonconverting rural hospitals (10.3 percent). Further, there was a positive association between length of time in the CAH program and estimated cost inefficiency. CAHs exhibited poorer values for a number of proxy measures for efficiency, including expenses per admission and labor productivity (full-time-equivalent employees per outpatient-adjusted admission). Non-CAH rural hospitals had a stronger correlation between cost inefficiency and operating margin than CAH facilities did.