The Affordable Care Act enacted significant Medicare payment reductions to providers, yet long-term effects of such major reductions on patient outcomes remain uncertain. Using the 1997 Balanced Budget Act (BBA) as an experiment, we compare long-run trends in 30-day readmission across hospitals with different amount of payment cuts. Using 100% Medicare claims between 1995 and 2011 and instrumental variable hospital fixed-effects regression models, we compared changes in 30-day readmission trends for 5 leading Medicare conditions between urban hospitals facing small, moderate, and large BBA payment reductions across 4 periods [1995-1997 (pre-BBA period), 1998-2000, 2001-2005, 2006-2001]. Patient sample includes Medicare patients who were admitted to general, acute, urban, short-stay hospitals in the United States 1995-2011. Sample size ranges from 1.4 million patients for acute myocardial infarction to 3 million for pneumonia. We found that 30-day readmission trends diverged post-BBA (2001-2005) between hospitals facing small and large payment cuts, where large-cut hospitals experience slower improvement in readmission rates relative to small-cut hospitals. The gap between small-cut and large-cut hospitals readmission trend was 6% for acute myocardial infarction, 4% for congestive heart failure and pneumonia (all P<0.01) in the 2001-2005 period. The gaps between hospitals were eliminated by the 2006-2011 period as the effect of BBA naturally dissipated over time. Although payment-cut differences are associated with widening gaps in readmission rates across hospitals, the negative association appears to dissipate in the long run.