SOME ARGUE THAT EQUALITY OF TREATMENT SHOULD BE THE OBJECT OF SENTENCING; THIS PAPER ARGUES FOR INEQUALITY OF SENTENCING JUST BECAUSE IT IS IMPOSSIBLE FOR THE LAW TO TAKE CIRCUMSTANCES INTO ACCOUNT. THE THEORY OF PARSIMONY, THE IDEA OF THE LEAST AFFLICTIVE SANCTION NECESSARY TO ACHIEVE A DEFINED SOCIAL PURPOSE, MANDATES INEQUALITY. CRACKDOWNS AGAINST DRUNKEN DRIVING TO REDUCE ROAD CARNAGE OR SEVERE ASSAULT SENTENCES FOR THOSE PARTICIPATING IN RACIAL VIOLENCE ARE EXAMPLES OF THE USE OF INEQUALITY TO DETER OTHERS FROM BEHAVIOR. TAX COURTS ARE VERY SELECTIVE IN CASES THEY PROSECUTE, BUT THE DETERRENT EFFECT CAN BE AS GREAT AS IF ALL TAX VIOLATORS WERE SENT TO PRISON, WHILE RESOURCES USED ARE MUCH LESS. THE NOTION OF REHABILITATION SHOULD HAVE NO BEARING ON SENTENCING: SENTENCES SHOULD BE PUNISHMENT. ANY PARTICIPATION IN REHABILITATION SHOULD BE VOLUNTARY. THE ABUSE OF PAROLE BOARDS USING REHABILITATION PROGRAM PARTICIPATION TO SET RELEASE DATES HAS BEEN WELL DOCUMENTED. THE THEORY OF DESERTS, THE AMOUNT OF PUNISHMENT A CRIME DESERVES, SHOULD SET MAXIMUM AND MINIMUM LIMITS FOR SENTENCES BUT CANNOT HANDLE THE FINE DISTINCTIONS IN BETWEEN. IT IS SUGGESTED THAT FLAT-TIME SENTENCES WITH TIME OFF FOR GOOD BEHAVIOR MAY MEET THE NEED FOR JUSTICE BUT THAT MANDATORY SENTENCING DOES NOT. THE AMERICAN LAW INSTITUTE'S MODEL PENAL CODE IS CONSIDERED A GOOD GUIDE TO REFORM IN SENTENCING PROCEDURES. IT IS SUGGESTED THAT MANDATORY SENTENCES WOULD INCREASE PLEA BARGAINING, SENTENCE BARGAINING, AND OTHER PRETRIAL MANEUVERS, THUS SUBSTITUTING PROSECUTORIAL DISCRETION FOR JUDICIAL DISCRETION.