Several previous studies have examined the correspondence between self-reports and the delayed identity match-to-sample performance they supposedly described. The present two experiments used similar procedures to explore different characteristics of the self-reports. In both studies, match-to-sample responses were successful (earned points) if they were both correct and faster than a time limit. Following each response, a computer-presented query asked whether the response had been successful, and subjects replied by pressing a "Yes" or "No" button. Experiment 1 analyzed self-report latencies from a previously-published study (Critchfield, 1993a). Latencies generally were longer for self-reports of failure than for selfreports of success. In Experiment 1, a "Yes" or "No" self-report was required to advance the session. In Experiment 2, self-reports were optional. In addition to "Yes" and "No" buttons, subjects could press a third button (a "nondiscdosure" option) to remove the self-report query without providing a "Yes" or "No" answer. Across a range of conditions, nondisclosures always occurred more frequently after match-to-sample failures than after successes (i.e., under conditions in which a self-report of failure would be appropriate). The effects observed in the two experiments are consistent with a history of differential punishment for uncomplimentary self-reports, which casual observation and some descriptive studies suggest is a common experience in United States culture. The research necessary to explore this notion should produce data that are of interest to psychologists both within and outside of Behavior Analysis.Skinner's (1957) Verbal Behavior may be most commonly viewed as an attempt to delineate relations unique to verbal behavior, but the analysis it contains rests firmly on the assumption that verbal relations reflect the same fundamental principles as other (primarily nonverbal) behavior that has been studied extensively in the laboratory. A thoroughgoing empirical evaluation of Skinner's analysis has been slow in coming. Much has been written about the relative scarcity of empirical investigations involving Skinner's "primary verbal operants" (e.g