Three sources of data were used: observations in trial courtrooms and during other stages of case processing (e.g., preliminary arraignment, screening, diversion), interviews with judges and attorneys, and dispositions resulting from several hundred pretrial decisions and over 600 courtroom hearings. Findings challenge the widespread assumptions that caseload pressures, cooperation among court participants, and process costs of litigation render informality and high guilty plea rates inevitable. In the municipal courts, roughly a fifth of all misdemeanor dispositions and a half of all determinations of guilt were the result of genuinely adversarial trials. Results suggest that an adversarial trial system need not impose prohibitive resources costs even when extended to misdemeanor cases involving little likelihood of imprisonment, and that process costs will not deter defendants from invoking formal procedures when the court culture is committed to providing trials. It is argued that an adversarial court organization offers substantially better prospects for preserving and enhancing those qualitative features of a court system that are the principal determinants of fairness and accuracy in the administration of justice. 18 tables and 267 footnotes. (Author abstract modified)