Researchers studying depression often encounter research participants in serious preexisting distress. Examining investigators' ethical responsibilities to these subjects, Stanton and New found that depression researchers reported actions that ranged from doing nothing to contacting both the distressed subject and a significant other. By experimentally manipulating consent form information regarding potential treatment referral, we examined whether subjects adjusted their responses on depression measures as a function of the level of follow-up they expected to receive. Results reveal that subjects who potentially could (...) receive the most intrusive intervention were less likely to report depressive symptoms than were subjects who anticipated less intrusive follow-up. Willingness to report depressive symptoms in particular conditions varied in part as a function of subject sex. Thus, ethical safeguards used in studies with subjects in preexisting distress may have consequences for the validity of self-report depression measures. (shrink)