With our growing dependence on clinical trials to assess the adequacy of available treatments for arthritic patients, attention is increasingly being focused on the tools we use in the design, analysis, and reporting of these studies and the ways in which we can improve them. This review outlines some of the major problems in terms of patient compliance, trial size, inadequate endpoints, and insufficient reporting, and highlights some of the work being done to overcome these difficulties. The greatest emphasis is placed on two issues. Firstly, there is a need for greater stringency in interpreting the results. Thus, more consideration needs to be devoted to the limitations of the measurements used and greater emphasis needs to be placed on analysis on the basis of "intention to treat". Secondly, there is a need to standardize common aspects of the trials, such as the detailed reporting of baseline data, a minimal set of outcome measures, comparable response thresholds, and the inclusion of statistics such as confidence intervals. Implementing these changes will increase comparability and yield a stronger possibility of significant results.