NO EMPIRICAL EVIDENCE EXISTSTO SUBSTANTIATE THE BELIEF THAT SMALL INSTITUTIONS ARE MORE EFFECTIVE THAN LARGE PRISONS IN REHABILITATING INMATES, DECREASING RECIDIVISM, OR PREVENTING PRISON VIOLENCE. FACTORS INFLUENCING THE EFFECTIVENESS OF VARIOUS PRISONS IN ACHIEVING CORRECTIONAL GOALS ARE APPARENTLY UNRELATED TO PRISON SIZE. THEREFORE, LARGE PRISONS WHICH ARE MUCH CHEAPER TO OPERATE SHOULD BE FAVORED OVER SMALL PRISONS. WHILE SMALL PRISONS MAY ENHANCE COMMUNICATIONS BETWEEN INMATES AND STAFF, NO EVIDENCE EXISTS THAT STAFF-INMATE RELATIONS ARE BETTER IN SMALL PRISONS THAN IN LARGE ONES; OR THAT CLOSE INMATE-STAFF RELATIONS NECESSARILY EFFECT SIGNIFICANT, POSITIVE, LASTING CHANGES IN INMATE BEHAVIOR. THE BUILDING OF LARGE PRISONS IN THE INTEREST OF COST SAVINGS DOES NOT INEVITABLY MEAN LOSING THE BENEFITS OF THE SMALL PRISON ENVIRONMENT. LARGE INSTITUTIONS CAN BE SO DESIGNED AS TO PROVIDE PRIVACY AND OTHER DESIREABLE CONDITIONS FOR HUMANE LIVING. ALSO, PRISONS WITH LARGE INMATE POPULATIONS CAN GENERALLY PROVIDE A BROADER RANGE OF PROGRAMS THAN COULD BE OFFERED IN A SMALL INSTITUTION. REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--RCB)