The research focused on the general tendency of judges to be more lenient or severe than their colleagues in felony sentencing. It did not focus on interaction effects between judges and particular characteristics of offenses and offenders. The research used a statistical model that aggregated the amount of inter-judge disparity across different cities. The model also tested whether any differences might be due to differences in caseloads arising from random case assignment and not due to differences in the judges’ sentencing decisions. Results of multivariate analyses revealed that the guidelines appeared to affect different crimes in different ways in different cities. The guidelines appeared on balance to be reducing aggregate inter-judge disparity and to be succeeding more often than failing, problem areas remained. Some data even suggested an increase in disparity in some places and with some types of offenses. Findings suggested that the sentencing guidelines are a qualified success. Reasons also exist to hope that additional improvements would further reduce sentencing disparity. Tables, review of earlier research, footnotes, and appended methodological information