At the time of the study, the U.S. Board of Paroles was required to consider three issues: (1) whether there is a substantial risk that the parole applicants will conform to parole conditions, (2) whether release would depreciate the seriousness of the inmate's crime or promote disrespect for the law, and (3) whether release would have a substantially adverse effect on institutional discipline. These requirements relate essentially to the probability of recidivism and the severity of the offense that occasioned incarceration. An actuarial table based on prior experience was constructed to estimate the probability of parole failure for cases in the Federal system. Prediction equations were then developed. The commisioners' rankings for the severity of offenses was then determined. It was possible to model the decisions of the board through scoring individual applicants on the dimensions of seriousness of offense (this determined the appropriate length of imprisonment according to offense severity) and probability of recidivism. The board eventually accepted the model as an explicit statement of its policy for determining the length of incarceration. To use the guidelines, the decisionmakers first ascertain the seriousness score of the offense. Next they find 11 items of information required to build up the prediction or salient factor score. They then calculate the score and identify the category (very high-low). They refer to the table of guidelines and read the intercept, which indicates the expected time to be served before release. Tabular data and 18 references are provided.