This article examines the theory and empirical methods of recent studies of law and litigation. It argues that the recent interest in longitudinal studies of trial court dockets proceeds from a deeply rooted functionalist theoretical tradition in empirical work on courts. Functionalist theory, through its sophisticated application in the work of James Willard Hurst, is described as the direct or indirect source of theory for longitudinal litigation studies. Though there are many reasons for suspecting that fuctionalist theory is inadequate, it has seldom been rejected through proper empirical testing of its hypotheses. The theory, often poorly conceptualized, is discussed here in detail. Hypotheses derived from a careful reading of Hurst and others are operationalized, employing data from the dockets of three West Virginia trial courts between 1870 and 1925, and a test of the hypotheses is constructed taking advantage of the historical variation in the economic development of three neighboring counties. Conclusions supported by the analysis of these data are largely prophylactic. Research can address the widespread problems of poor conceptualization and data analysis that have limited the value of much longitudinal trial court docket research. If court function is to continue to be the focus, theory, it is argued, must take account of the social organization of relations between litigants outside the courts, as well as the organization of the courts themselves. Data analysis must address the measurement of the power, litigation capacity, and perception of usefulness of litigation held by litigants. Illustrating the value of hypothesis testing as employed here, it is contended, these conclusions about appropriate future directions and methods for longitudinal court docket research are based on properly supported findings.