This study investigates the effect of a mental content of presented stimuli, normal aging and individual differences in cognitive abilities on temporal limits of an integration mechanism. Younger and older subjects grouped together the beats generated by a metronome. Subjects were asked to listen to the beats of a metronome and to accentuate mentally every second, third, fourth...etc. beats, to create a subjective rhythm. This rhythm exists, in fact, only in subjects' mind and not objectively. Subjects reported verbally how many clicks they were able to integrate into a perceptual unit. On this basis, the time interval during which subjects were able to integrate temporally separated stimuli was calculated (number of beats reported as being integrated x time distance between beats) for different metronome frequencies. The results show, firstly, that the length of integration periods significantly depends on the frequency of presented metronome beats. When the frequency of metronome beats is high, the time interval during which the subjects' integrate beats into a single perceptual unit is shorter. Secondly, older adults integrate information during a longer time interval then the younger ones. Thirdly, the length of an integration period is related to a subjects' level of cognitive ability. These results suggest that the length of an integration period is not a constant, stable feature, but varies across the life span depending on the mental content of the information presented and individual factors.