My dissertation is based differentiating individually experienced, musical time, and shared, astronomical time. I combine the knowledge gained in cognitive science experiments with music-theoretical approach following the path from sound received to the hierarchized representation of the temporal component of music. My focus are those possible mechanisms that allow deviation of musical time from the astronomical one and the factors influencing them. As concerns music theory, I chose the analytical tools provided by the General Theory of Segmentation and the Recontextualization Theory by D. A. Hanninen. The inner sharing of pulse by the listener or performer with the pulse of sounding music I consider to be the decisive moment for/in the deviation of musical time. I link the pulse boundary with prospective and retrospective strategies of the perceiving person. My considerations unfold on an axis between two poles of pulse and gesture, i.e. between full kinaesthetic synchronization with the outer, shared time and the maximum deviation dependent on the individual manner the gesture is internally represented. I convey the hypothesis that the deviation of musical time is possible only with retrospective strategies of processing of musical events. My theoretical considerations and suitability of the chosen analytical tools/means are verified in the second part, in analyses of compositions by G. Grisey, O. Messiaen, P. Kotík, P. Graham and A. Breier. Enclosed are also some theoretical reflections of these composers concertning their work with time.