MULTIVARIATE ANALYSIS WAS USED TO TEST FOUR KINDS OF VARIABLES: (1) COVERAGE VARIABLES (I.E. COLUMN INCHES OF SPACE, FREQUENCY OF FRONT PAGE COVERAGE); (2) POLICY SIGNIFICANCE (I.E., FREQUENCY OF SUBSEQUENT CASE CITATIONS BY COURTS AND LAW JOURNALS); (3) CASE CHARACTERISTICS (I.E., CIVIL VERSUS CRIMINAL, PRESENCE OF AMICI CURIAE PARTIES); AND (4) DECISION CHARACTERISTICS (I.E., LENGTH OF THE MAJORITY OPINION AND THE NUMBER OF DISSENTING VOTES). THIRTY-FIVE HYPOTHESES WERE TESTED. THE 10 NEWSPAPERS PUBLISHED 277 ACCOUNTS OF THE 139 DECISIONS, FOR A MEAN REPORTING PERFORMANCE OF 20 PERCENT. SOME 55 PERCENT OF THE DECISIONS WERE IGNORED BY ALL 10 PAPERS; 37 PERCENT WERE COVERED BY 1 TO 6 PAPERS; AND 8 PERCENT WERE COVERED BY 7 OR MORE PAPERS. THE TWO METROPOLITAN NEWSPAPERS PROVIDED THE MOST COMPREHENSIVE COVERAGE. DISCRIMINANT ANALYSIS INDICATED THAT CASE CHARACTERISTICS WERE THE WEAKEST PREDICTORS AND DECISION CHARACTERISTICS THE STRONGEST PREDICTORS OF COVERAGE. IN CONTRAST, MULTIPLE REGRESSION ANALYSIS INDICATED THAT POLICY SIGNIFICANCE VARIABLES WERE THE STRONGEST GROUP OF PREDICTORS. RESULTS INDICATED THAT NEWSPAPERS FALL SHORT OF THE REQUIREMENTS OF THEIR WATCHDOG ROLE. WITH RESPECT TO CASE AND DECISION CHARACTERISTICS, NEWSPAPER REPORTING EITHER RESEMBLES RANDOM BEHAVIOR OR IS INFLUENCED BY FACTORS NOT INCLUDED IN THIS STUDY. FURTHER RESEARCH USING MORE VARIABLES AND MORE NEWSPAPERS IS RECOMMENDED. A LITERATURE REVIEW, ADDITIONAL RESULTS AND CONCLUSIONS, FOOTNOTES, TABLES, A REFERENCE LIST, AND APPENDIXES PRESENTING STUDY METHODOLOGY AND ADDITIONAL RESULTS ARE INCLUDED. (CFW)