Sounds of human laughter compose quite effectual stimuli that usually facilitate positive responses. We have studied the mechanisms of such effects and investigated how changes in particular acoustical signal parameters affect the evaluation of laughter. Effects were assessed by evaluating self-report data of human subjects who had been exposed to playbacks of experimentally modified laughter material and, for control, also to samples of natural laughter. The modified laughter phrases were generated by first analysing samples of natural laughter, and then using these data to synthesise new laughter material. Analyses of subjects' responses revealed that not only samples that resembled the rhythm of natural laughter (repetition interval of about 0.2 s) were evaluated positively. Instead we found that series with a wide range of repetition intervals were perceived as laughter. The mode of parameter changes within the model series had an additional clear effect on the rating of a given playback sample. Thus, an intra-serial variation of rhythm or pitch received ratings that were closer to ratings of natural laughter (control) than did a stereotyped patterning of stimuli. Especially stimuli with decreases in pitch were well suited to elicit positive reactions. In conclusion, our results showed that features of parameter variations can make human laughter particularly effectual.