Four experimental interview styles, each recommended by experts in the field, were compared for their efficiency in eliciting factual information during the initial diagnostic interviews with the mothers of children referred to a psychiatric out-patient clinic. If encouraged to talk freely, mothers tended to mention most (but not all) key issues without the need for standardized questioning on a pre-determined range of topics. However, systematic questioning was essential in order to obtain good quality factual data. Better data were obtained when interviewers were sensitive and alert to factual cues and chose their probes with care. Clinically significant factual information, idiosyncratic to the family and outside the range of standard enquiry was common, but was obtained satisfactorily with all four styles. No one style was generally preferred by informants. The advantages of systematic questioning for obtaining factual information were not associated with any disadvantages with respect to the eliciting of emotions and feelings.