This article addresses the role of response latencies in affecting the attitude-behavior consistency in a German subnational referendum. As voters faced a comparatively easy choice in this referendum, it puts the hypothesis concerning the role of attitude accessibility in increasing attitude-behavior consistency to a particularly hard test. Utilizing data from a two-wave panel survey, the analysis examines the effect of response latencies on the attitude-behavior consistency concerning participation and vote choice. The evidence confirms hypotheses derived from attitude-consistency theory only in a limited number of cases. The institutional setting and the nature of choice thus appear to make a difference. Moreover, substantive findings depend partially upon the procedure to measure response latencies. Accordingly, sensitivity tests should be employed by default. Irrespective of operationalization, response latencies play a crucial role when it comes to respondents who answered that they would “perhaps” participate. Whereas a quickly uttered “perhaps” was indicative of a rather low likelihood of participation, a slowly given “perhaps” indicated a considerably high probability.