Initiated in 1985, the study examined existing data and gathered additional information on 50 felony offenders and 60 driving-under-the-influence cases. Data were also gathered through structured interviews with policymakers and program managers. The current population projections for both the jail and honor farm were found to be inaccurate, due to the inappropriate use of demographic and inmate data. Defendants with warrants and holds and those unable to secure commercial bail or release on recognizance constituted the two main pretrial groups that increased the jail population. Increases in the use of jail sentences and in the length of sentences for drunk driving have strongly affected the honor farm population. Alternatives that would reduce these populations would be a supervised pretrial release program and a treatment and sentence modification program for drunk drivers. A more sophisticated projection model is also needed for predicting inmate populations. Included are data tables and appendixes presenting a list of the individuals interviewed and a description of the projection model.