Relationships between attitude and overt behaviors were examined in two studies. Attitude toward abortion was assessed by questionnaire in undergraduate classes. Two to three months later pro- and anti-abortion subjects were seen individually, for an ostensibly unrelated project, and were asked to transcribe resolutely pro- or anti-abortion scripts. Supposedly the recordings would be used in a future study designed to convince hospital personnel either to support or refuse to participate in abortion procedures. In Study I, where respondents were allowed to transcribe either script, a strong attitude / choice-behavior correlation resulted: ϕ = .60. In Study II, the script was assigned randomly, creating attitude-congruent and attitude-discrepant conditions. Three criteria were affected in the expected way by attitude-congruency / discrepancy: (a) recording one's name with the persuasive message (ϕ = .56); (b) authorizing use of an identified tape (ϕ =.44); and (c) self-estimates of persuasiveness (ϕ = .36). Two other intuitively relevant behaviors werenotsignificantly associated with attitude-congruency /discrepancy. The observed correlations were high, compared to previous attitude-behavior research, and some possible explanations are offered.