The effects of two experimental interview styles, designed to differ in the extent of their use of active feeling-oriented techniques but similar in their use of active fact-oriented techniques, were compared in initial diagnostic interviews with the mothers of children referred to a psychiatric clinic. The style that employed a higher level of actively responsive feeling-oriented techniques elicited more emotional expression and more often obtained certain feelings of potential diagnostic significance. The actively responsive style was more effective in increasing the amount of feeling expressed if mothers' spontaneous rate of expression was relatively low.