2 studies investigated the relationship between the expressive behaviors of smiling and laughing (mirth) and the evaluations of the funniness of cartoon stimuli. Study 1 replicated past findings that suggest that feedback from mirth reactions directly influences funniness ratings in female Ss but does not directly influence funniness ratings in male Ss. Study 2 provided evidence of a sex difference in the link between mirth and evaluation. It was hypothesized that making Ss aware of their mirth, by asking them to self-observe and rate their own smiling and laughing, isolates the mirth reactions and reduces their influence upon evaluations of funniness. The elimination of the influence of mirth should lower funniness ratings of female Ss and eliminate any increase in rated funniness produced by canned laughter. No such effects were predicted for males. The data support the hypotheses. A wider range of possible causes for self-observation effects is given than is mentioned in most recent theories of self-observation. (31 ref) (PsycINFO Database Record (c) 2016 APA, all rights reserved)