The concepts of the "therapeutic alliance" (Zetzel, 1956) and "working alliance" (Greeson, 1965) are traced in their antecedents to Freud and other analysts and compared especially with Freud's "analytic pact." Differences both in theory and practice are elucidated with the aid of case material. While the therapeutic and working alliances are often used interchangeably, they are found to take up their positions at diametrically opposite points along a continuum defined by the analytic pact and show a marked tendency to depart from the guidance offered the traditional analysis by the fundamental rule. They should be regarded more as exercises in analytically oriented psychotherapy than parameters of the traditional technique. Nevertheless, as a basis for comparisons, it should be recognized that the latter technique: (1) does not formulate in theory many of the measures left for pragmatic fulfillment; (2) does not include genetic, structural and adaptive viewpoints that later developments require. Current tendencies to include the analyst's self-observations in relation to the total analytic process point up the growing influence of these more recent considerations.