Although the application of scientific research methods to policy issues dates back to world War II, the application of social science research to court reform is relatively new. This suggests that the problem is not misuse of the research, but nonuse. Four reasons for nonuse are social science's failure to address key policy issues, the problem of timeliness with social data, irrelevance of research to user concerns, and the prevalence of bad research. Increasingly, court reform research involves the user of fashionable but inappropriate tools of analysis or modeling, tools largely discredited as inapplicable to policy research in other fields. A few of these are queuing theory, optimum sequencing models, Markov processes, regression and correlation, and analysis of variance. Court administrators could apply social research techniques in a useful way, however. Equal employment data could be analyzed to establish hiring practices. Cost-of-living and income statistics could be used in setting salary scales. Weighted caseload statistics could be used as a basis for assigning judges and nonjudicial staff.