This study was focused on the effect of voting split and petitioning parties on voting behavior, using a systematic model to predict voting patterns for five Supreme Court justices serving across the decade 1947-56. The model is theory-integrating in that it combines propositions from attitudinal and role theories in a logically coherent fashion. In general it posits that liberal and conservative justices are predisposed in review decisions to vote for underdogs and upperdogs, respectively, but that a willingness to act on such predispositions will be tempered by the role prescription of impartiality. The model suggests this conflict will be resolved to some extent via the impartiality while actually voting according to a bias. The model predictions were reliable in each case. The study inferred that liberal and conservative Supreme Court justices are not only motivated to vote their predispositions toward litigants seeking review but also to minimize the costs of such acts by nesting them in conditions under which their significance is not fully appreciated. The limitations of such findings are, however, that only five justices were studied across a single decade. Nevertheless, assuming that the model has some validity, and given that the role prescription of impartiality permeates the political, legal, and governmental systems of the U.S., research designed to identify and measure response to such a consideration throughout official decisionmaking agencies should be encouraged by these results. That is, are justices in other courts and decisionmakers in nonjudicial arenas prone to such tendencies in the decisional settings in which they operate? About 20 references and several tables are provided. (Author abstract modified)