This study compared rectal temperature rhythms in groups of subjects who either did (n = 5) or did not (n = 7) show a clear postlunch dip in performance at a monotonous (25-30 min) vigilance task. Performance was tested every 2h in a standardized routine with lunch replaced by hourly liquid food supplements. Those showing the postlunch performance dip had a higher amplitude and later peaking 12h component to their rectal temperature rhythm that those who did not, resulting in flat, rather than rising, temperatures over the 10:00-15:00 time interval. The effect could not be explained by intergroup differences in prior sleep, morningness, or gender, although there was a trend (p = 0.09) for the "dip" group to be slightly younger (21.8y vs. 24.2y). The postlunch dip appears to be an endogenous phenomenon individually determined, but related to the strength of the (12h) harmonic of the circadian system.