GRIEVANCES AND RESTLESSNESS AMONG PRISONERS LED TO LEGISLATION, IN 1943 IN MASSACHUSETTS AND IN 1957 IN CONNECTICUT, ESTABLISHING A SENTENCE REVIEW DIVISION CONSISTING OF THREE TRIAL JUDGES FROM THE COURTS OF FIRST INSTANCE. ALL OFFENDERS SENT TO THE PRISON MAY REQUEST REVIEW. THE CONNECTICUT REVIEW DIVISION IS REQUIRED TO PROVIDE WRITTEN REASONS FOR ITS DECISIONS; THE MASSACHUSETTS DIVISION IS NOT. THE BOARDS MAY REDUCE OR INCREASE A SENTENCE. THE FREQUENCY OF SENTENCE APPEALS VARIES WITH THE SEVERITY OF THE ORIGINAL SENTENCE AND DEPENDING UPON WHETHER THE SENTENCE WAS IMPOSED AFTER A GUILTY PLEA OR AFTER TRIAL. SENTENCES UP TO 5 YEARS IN MASSACHUSETTS AND UP TO 3 YEARS IN CONNECTICUT RARELY ARE MODIFIED AND HENCE SELDOM ARE APPEALED. OVERALL MODIFICATION RATES ARE LOW IN BOTH STATES. OF ALL ELIGIBLE SENTENCES, MASSACHUSETTS REDUCED 2.3 PERCENT AND INCREASED 0.2 PERCENT, WHILE CONNECTICUT REDUCED 1.2 PERCENT AND INCREASED NONE. MEASURES OF SENTENCE DISPARITY HAVE BEEN DEVELOPED FOR SOME FEDERAL COURTS, BUT NOT FOR STATE COURTS. IF DISPARITY IN STATE COURTS IS ANYWHERE NEAR THAT FOUND FOR FEDERAL COURTS, THE REDUCTIONS BY THE MASSACHUSETTS AND CONNECTICUT REVIEW BOARDS ARE MINIMAL EFFORTS. IN BOTH STATES, THE REASON MOST FREQUENTLY CITED FOR REDUCING A SENTENCE IS THE BOARD'S PERCEPTION THAT THE SENTENCE IS OUT OF LINE WITH GENERAL PRACTICES. THE WRITTEN OPINIONS REQUIRED OF THE CONNECTICUT BOARD GENERALLY FAIL TO ARTICULATE REASONS FOR REJECTING OR ALLOWING AN APPEAL. THE STUDY INDICATES THAT A POTENTIAL INDIRECT EFFECT OF THE BOARD -- THAT OF CONTRIBUTING TO A MORE GENERAL REDUCTION OF SENTENCE DISPARITY BY ALLOWING TRIAL JUDGES TO SEE AND COMPLY WITH THE BOARD'S STANDARDS -- IS MINIMAL. SUGGESTIONS FOR DEVELOPING SENTENCING GUIDELINES ARE OFFERED. SUPPORTING DATA ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED--LKM)