Interviews with more than 2,000 adults showed that the presence of third parties (including parents) during the interview significantly influenced respondents' willingness to reveal illicit drug use. Among married respondents, presence of a spouse resulted in higher reporting of illicit drug use, while the presence of adults other than the spouse had a consistent negative effect on drug use reports. The pattern of findings suggests that the direction of effects of third party presence is linked to two factors: the extent of the third party's knowledge of the information requested, and the degree of the third party's stake in the respondent's answers. Tests of interactions between privacy and interview mode failed to support the hypothesis that the use of self-administered answer sheets reduces privacy effects compared with interviewer-administered interviews. Tables, references