Data were collected on 520 Federal probation cases processed in the Northern District of Alabama. Independent variables of the study were levels of supervision recommended by the risk-screening instruments and probation officers; the dependent variable was the probation outcome, dichotomized and coded as success or failure. The results showed that probation officers' assessment of offenders, if used as a predictive scale, appears to be very efficient. Almost all of the offenders classified as requiring minimum supervision successfully completed their probation terms, but only 52 percent of those classified as requiring maximum supervision successfully completed probation. The results suggest that probation officers can benefit from the use of risk-screening instruments. In particular, use of the USDC75 by officers as additional information in cases which fall near the boundaries of classification categories might allow officers to increase the number of offenders in maximum supervision categories. 2 tables, 19 notes, and 4 appendixes