STUDY METHODS INCLUDED REVIEWS OF COURT DOCKET SHEETS, ANALYSIS OF COURT FILES FOR 59 NONRANDOMLY SELECTED CASES AND ADDITIONAL FILES FOR A SUBSAMPLE OF 38 CASES. IN ADDITION, THE JUDGE HANDLING THE CASES, WELFARE DEPARTMENT EMPLOYEES, AND OTHERS WERE INTERVIEWED. UNDER ILLINOIS LAW, WELFARE RECIPIENTS OR APPLICANTS CONVICTED OF FRAUD MAY BE FINED UP TO $1000 OR IMPRISONED FOR UP TO ONE YEAR. CASE FLOW WAS ANALYZED IN DETAIL, INCLUDING DISCOVERY, ANALYSIS, PROSECUTION, COURT TRIAL, AND DISPOSITION. THE MAIN FACTOR CONSIDERED IN THE DECISION TO PROSECUTE WAS THE RECIPIENT'S INTENT, WHICH WAS USUALLY DETERMINED BY SUBJECTIVE ANALYSIS. DEFENDANTS RARELY USED ATTORNEYS. OVER 70 PERCENT OF THOSE ENTERING PLEAS PLEADED GUILTY. ONE QUARTER OF THOSE PLEADING NOT GUILTY WERE FOUND NOT GUILTY. THE SENTENCE WAS ALMOST ALWAYS PROBATION FOR SIX MONTHS TO FIVE YEARS. ONE YEAR WAS THE MOST COMMON TERM. NO SINGLE RULE GOVERNED THE LENGTH OF PROBATION IMPOSED. IN OVER HALF THE CASES, THE DEFENDANT WAS ORDERED TO MAKE RESTITUTION AS A CONDITION OF PROBATION. A SMALL PERCENTAGE OF CASES RESULTED IN A DEFENDANT'S PLACEMENT UNDER SUPERVISION. IT IS RECOMMENDED THAT (1) DEFENSE ATTORNEYS BE PROVIDED OR CASES TRANSFERRED TO ANOTHER COURT BRANCH, (2) RESTITUTION NOT BE MADE A CONDITION OF PROBATION FOR ACTIVE WELFARE CASES, (3) SENTENCES BE REVIEWED ANNUALLY, (4) AUDITORS OFFICE GIVE MORE INFORMATION TO SUSPECTS BEFORE AND DURING INTERVIEWS, AND (5) CRIMINAL CONVICTIONS FOR WELFARE FRAUD BE REPLACED WITH USE OF CIVIL REMEDIES. A DIAGRAM, FIGURES, TABLES, AND FOOTNOTES WHICH INCLUDE REFERENCES ARE INCLUDED.