CRIME STATISTICS HAVE LONG BEEN A TOPIC OF CONCERN TO CRIMINOLOGISTS. AS EARLY 1778 BENTHAM URGED THE PUBLICATION OF DATA ON PRISONERS AS AN AID TO LAWMAKERS. IN 1829, GUERRY PUBLISHED A PAMPHLET OF CRIME STATISTICS, AND HIS TREATISE OF 1833, 'ESSAYS CONCERNING THE MORAL STATISTICS IN FRANCE,' IS CREDITED WITH ORIGINATING THE SCIENCE OF CRIMINAL STATISTICS. QUETELET, A CONTEMPORARY OF QUERRY, PROVIDED DISCUSSIONS OF ASSUMPTIONS UNDERLYING THE UTILIZATION OF OFFICIAL CRIME DATA AND THEIR DEFICIENCIES. THE FIRST ATTEMPT TO GATHER CRIME DATA IN THE U.S. WAS IN 1834 BY THE STATE OF MASSACHUSETTS. THE FEDERAL GOVERNMENT BEGAN TO COLLECT CRIME DATA IN 1850 IN CONJUNCTION WITH THE CENSUS, AND SUCH COLLECTION WAS CONTINUED IN SUBSEQUENT CENSUSES. IN 1880, CRIMINAL STATISTICS INCLUDED DATA FROM POLICE REPORTS, AND IN 1904, THE FOCUS OF INQUIRY WAS EXPANDED TO INCLUDE COMMITMENTS TO PENAL INSTITUTIONS, PERSONS IN PRISONS ON A GIVEN DAY, AND A MEASURE OF GEOGRAPHIC MOVEMENT OF THE PRISON POPULATIONS DURING THE YEAR. THE FIRST COLLECTION OF JUDICIAL STATISTICS ON A FEDERAL LEVEL BEGAN IN 1906. CRIMINAL STATISTIC COLLECTION WAS SPORADIC FROM 1908 TO 1925; CENSUS DATA FROM 1926 THROUGH 1931 AGAIN INCLUDED PRISONER DATA. THE U.S. FEDERAL BUREAU OF PRISONS, BEGUN IN 1929, AND THE INTERNATIONAL ASSOCIATION OF CHIEFS OF POLICE (IACP) BOTH CONTRIBUTED TO THE FURTHER EVOLUTION OF CRIME STATISTICS COLLECTION. IN 1930, THE U.S. CONGRESS TRANSFERRED THE RESPONSIBILITY FOR UNIFORM CRIME REPORTING FROM IACP TO THE FEDERAL BUREAU OF INVESTIGATION, WHERE IT STILL REMAINS. THE 1931 NATIONAL COMMISSION ON LAW OBSERVANCE AND ENFORCEMENT REPRESENTED THE FIRST GOVERNMENTAL EVALUATION OF THE ENTIRE CRIMINAL JUSTICE SYSTEM AT ONE TIME. IT DISCUSSED POLICE, PRISON, COURT, PROBATION, AND PAROLE STATISTICS, AND RECOMMENDED THE CENTRALIZATION OF STATISTICS COLLECTION, THE ADOPTION OF UNIFORM COLLECTION PROCEDURES, AND THE EXPANSION OF THE NUMBER OF REPORTING AGENCIES. IN 1957, CRESSEY MADE A CRITICAL EXAMINATION OF POLICE STATISTICS, AND CIPES (1968) AND SEIDMUN AND COUZINS (1974) HAVE CRITIQUED POLITICAL ASPECTS OF CRIME REPORTING. OTHERS HAVE CRITICALLY EXAMINED PROCEDURAL PROBLEMS IN CRIME STATISTICS COLLECTION. A NUMBER OF CRITICISMS HAVE ALSO BEEN AIMED AT THE UNIFORM CRIME REPORTS. THERE HAVE BEEN FEW EMPIRICAL ATTEMPTS TO EQUATE OFFICIAL STATISTICS WITH ACTUAL INCIDENCE, BUT HINDELANG (1974) AND SEVERAL OTHERS HAVE FOUND THAT POLICE DATA MAY BE MORE ACCURATE INDICATORS OF ACTUAL CRIME THAN PREVIOUSLY THOUGHT. IT IS CONCLUDED THAT THE UNIFORM CRIME REPORTS ARE A VALUABLE TOOL FOR ACADEMIC AND POLICY-RELATED INQUIRY IF USED WITH ADEQUATE CAUTION. NOTES ARE INCLUDED. (JAP)