DEMOGRAPHIC FACTORS, OFFENSE BACKGROUND, LENGTH, STAGE, AND NATURE OF CONFINEMENT FOR THE EXPERIMENTAL GROUP ARE COMPARED, WHERE POSSIBLE, WITH THOSE FOR A SAMPLE OF 1,383 INMATES IN 4 MAXIMUM-SECURITY PENITENTIARIES IN 1977. WHEN THE DATA ARE COMPARED WITH GENERAL STATISTICS, FINDINGS SHOW THAT (1) THERE IS A MUCH HIGHER RATE OF SUICIDE AMONG THE INMATES THAN AMONG THE GENERAL POPULATION, (2) THERE IS A HIGHER RATE OF SUICIDE AMONG YOUNGER, UNMARRIED MALES, (3) MOST SUICIDES HANG THEMSELVES WITH STANDARD ISSUE MATERIAL SUCH AS CLOTHING, AND (4) SUICIDES ARE PARTICULARLY LIKELY TO OCCUR AMONG THOSE INMATES SERVING SHORT OR LIFE-TERM SENTENCES, ESPECIALLY DURING THE FIRST 12 MONTHS OF CONFINEMENT, AMONG THOSE PERSONS OUTSIDE OF GENERAL POPULATION CONDITIONS SUCH AS IN PSYCHIATRIC WINGS. THERE IS NO APPARENT DIFFERENCE IN SUICIDE-PRONENESS BETWEEN PRISONERS INCARCERATED FOR VIOLENT OFFENSES AND THOSE IMPRISONED FOR PROPERTY OFFENSES. EVIDENCE DOES SHOW THAT MANY INMATES SUICIDES HAD EXTENSIVE PSYCHIATRIC HISTORIES INCLUDING PREVIOUS THREATS OF, OR ATTEMPTS AT, SUICIDE. EFFECTIVE SUICIDE PREVENTION POLICIES SHOULD INCLUDE SWIFTER EMERGENCY RESPONSES; FREQUENT SURVEILLANCE OF INMATES SHOWING SUICIDAL INTENTIONS; LESS RELIANCE UPON PUNITIVE RESPONSES; AND GREATER INVOLVEMENT OF CLINICAL STAFF, PARAPROFESSIONALS, AND AREA KIN. A SUBSTANTIAL RESEARCH AND DEVELOPMENT EFFORT SHOULD BE DIRECTED TOWARDS SUICIDAL BEHAVIOR IN JAILS, PRISONS, AND PENITENTIARIES. AN APPENDIX AND TWO BIBLIOGRAPHIES, ONE WITH ANNOTATIONS, ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--AOP)