THE LONGITUDINAL METHOD HAS BEEN USED TO INVESTIGATE CRIMINAL CAREERS, ESPECIALLY THE INCIDENCE AND PREVALENCE OF OFFICIAL DELINQUENCY AT DIFFERENT AGES, THE PEAK AGE FOR CONVICTIONS, THE RELATIONSHIP BETWEEN JUVENILE DELINQUENCY AND ADULT CRIME, AND OFFENSE SPECIALIZATION. IT HAS ALSO BEEN USED TO PREDICT THE ONSET OF CONVICTIONS, RECIDIVISM, AND THE ENDING OF CRIMINAL CAREERS; TO STUDY THE EFFECTS OF PENAL TREATMENTS AND OTHER EVENTS SUCH AS MARRIAGE ON DELINQUENCY; AND TO INVESTIGATE THE TRANSMISSION OF CRIMINALITY FROM ONE GENERATION TO THE NEXT. LONGITUDINAL AND CROSS-SECTION METHODS EACH HAVE A PART TO PLAY IN RESEARCH INTO CRIME AND DELINQUENCY AND EACH HAS ADVANTAGES AND DISADVANTAGES. MAJOR METHODOLOGICAL QUESTIONS CONCERN THE RELATIVE VALUE OF PROSPECTIVE VERSUS RETROSPECTIVE RESEARCH AND OF INTERVIEWS VERSUS OFFICIAL RECORDS. PRACTICAL PROBLEMS OF LONGITUDINAL SURVEYS INCLUDE FUNDING, STAFFING, ATTRITION, AND ETHICAL ISSUES. MORE PROSPECTIVE LONGITUDINAL RESEARCH PROJECTS, INCLUDING INTERVIEWS, SHOULD BE UNDERTAKEN TO PROVIDE BASIC INFORMATION ABOUT THE NATURAL HISTORY OF DELINQUENT BEHAVIOR AND THE PROGRESSION OF DETECTED OFFENDERS THROUGH THE CRIMINAL JUSTICE SYSTEM. AN APPENDIX WITH A SELECTION OF MAJOR LONGITUDINAL SURVEYS AND A LIST OF REFERENCES ARE INCLUDED. (AUTHOR ABSTRACT MODIFIED - MJW)