Readmission rates are being proposed as an outcome indicator of hospital-level quality despite the lack of evidence of a relationship between a readmission and the quality of care provided during the preceding hospital stay. This study examined this relationship by comparing the quality of care provided to samples of 134 unplanned readmissions and 158 nonreadmissions from 50 Department of Veterans Affairs medical centers. Four groups of commonly used quality indicators and seven readmission risk factors were included in a logistic regression analysis. Inclusion of the quality indicators significantly increased the predictive power of the logistic models, with "unexpected transfer to a special care unit" being the only significant variable.