DECIDING HOW LONG IMPRISONMENT SHOULD BE FOR A PARTICULAR OFFENDER CAN BE RATIONAL IN AN ECONOMIC SENSE ONLY IF THE FOLLOWING FACTORS CAN BE CALCULATED: NET COST TO THE COMMUNITY OF INCARCERATION FOR A SPECIFIED PERIOD OF TIME RELATIVE TO INCARCERATION FOR A GREATER OR LESSER PERIOD; EFFECT UPON THE FUTURE BEHAVIOR OF PRISONERS OF THE IMPRISONMENT EXPERIENCE; AND DETERRENT EFFECT OF IMPRISONMENT ON THE BEHAVIOR OF OTHERS. CRIMINOLOGY LITERATURE INDICATES THAT NONE OF THESE CALCULATIONS ARE POSSIBLE. THERE IS SOME EVIDENCE THAT WHAT DISTINGUISHES ENGLAND AND WALES FROM OTHER EUROPEAN COUNTRIES WITH SMALLER PRISON POPULATIONS IS THE GREATER LENGTH OF SENTENCES PASSED BY THE COURTS. THE JUDICIAL DECISION AS TO SENTENCE LENGTH IS, THEREFORE, CRITICAL IN DETERMINING THE SIZE OF PRISON POPULATIONS. USE OF THE TIME DIMENSION IN SENTENCING IS DISCUSSED, WITH CALCULATION PROCEDURES AND SUPPORTING TABULAR DATA PROVIDED. SENTENCES CLUSTER AROUND WHOLE TIME UNITS, AND GAPS BETWEEN CLUSTERS LENGTHEN AS SENTENCE LENGTH INCREASES. REFERENCES ARE INCLUDED. (DEP)