To replicate and extend Dozier's (1992) test of Tomm's hypothesis about the differential effects of questioning styles on therapeutic alliance, an analogue study was conducted. Twenty-eight family triads, each including a son and his parents, viewed four videotaped, simulated family therapy scenarios in which Tomm's four questioning styles were separately portrayed. Participants were asked to identify with the client whose role corresponded to theirs (that is, father, mother, or son) and, on the basis of this, to rate the client's alliance with the therapist. They were also asked to rate the overall alliance between the family and the therapist. Finally, having viewed all four scenarios, they were invited to rate comparatively the quality of the therapeutic alliance across the four questioning styles. Compared with strategic and lineal questioning styles, circular and reflexive questions led to higher ratings of therapeutic alliance on all three measures. The results of this study support Tomm's hypothesis that questioning styles based on circular assumptions lead to a better therapeutic alliance at an individual and systemic level than do questions based on lineal assumptions.