THE STUDY SUBJECTS WERE 551 BURGLARS PAROLED DURING THE 1967-68 FISCAL YEAR. THE BENEFITS OF THE PAROLE SYSTEM WERE VIEWED IN TERMS OF THE ADDED PRODUCTION AND REDUCED INCARCERATION COSTS RESULTING FROM THE EARLY RELEASE OF INMATES, WHILE THE COSTS WERE VIEWED IN TERMS OF THE ADMINISTRATION OF THE PAROLE SYSTEM, THE CRIMES COMMITTED BY PAROLEES, THE CRIMES INDUCED BY THE LOWERED PENALTY WHICH RESULTS FROM PAROLE, AND THE REDUCED IN-PRISON PRODUCTION. AN ADDED POSSIBLE BENEFIT OR COST--THE DIFFERENTIAL PROBABILITY OF RECIDIVISM BETWEEN PAROLEES AND MANDATORY RELEASE --WAS DISCUSSED BUT NOT MEASURED BECAUSE OF A LACK OF DATA. WHEN INSUFFICIENT DATA PRECLUDED A DIRECT MEASUREMENT OF ONE OF THE FACTORS IN THE COST-BENEFIT ANALYSIS, INTERVAL ESTIMATES WERE CALCULATED. IN EACH CASE, ASSUMPTIONS WERE MADE WHICH SERVED TO MAXIMIZE BOTH BENEFITS AND COSTS, THEREBY ENABLING AN INTERVAL ESTIMATE OF THE FINAL COST-BENEFIT RATIO TO BE OBTAINED. THE NET BENEFIT RESULTING FROM THE PAROLE OF BURGLARS IN PENNSYLVANIA RANGED FROM MINUS $2,216,887 TO PLUS $1,894,346. SOME LIMITATIONS IN THE DATA EMPLOYED ARE NOTED. IT IS CONCLUDED THAT THE INTERVAL ESTIMATE OBTAINED DOES NOT PERMIT A JUDGMENT AS TO WHETHER OR NOT THE PAROLE OF BURGLARS IN PENNSYLVANIA IS ECONOMICALLY JUSTIFIED; STILL, IT IS BELIEVED THE STUDY IS OF VALUE IN PROVIDING SOME MEASURE OF THE MAXIMUM LOSS OR GAIN WHICH SOCIETY WILL ENCOUNTER IN THE PAROLE OF BURGLARS. DATA ARE PROVIDED AND A BIBLIOGRAPHY IS INCLUDED. (RCB)