Under the influence of artificial zeitgebers, human circadian rhythms can be entrained only within limited ranges of periods; different overt rhythms may show different entrainment limits. When the period of a zeitgeber is varied slowly but continuously, entrainment limits can be evaluated precisely. An overt rhythm is synchronized to the zeitgeber only up to a certain day, or period respectively, until it breaks away from the zeitgeber and starts to freerun. The interindividual comparison among different subjects shows that the range of entrainment is positioned nearly symmetrically around the freerunning period. Its width depends strongly on the freerunning period; it increases with lengthening freerunning period. As the consequence, subjects with a freerunning period only slightly shorter than 23 h would fail to become synchronized to the natural 24-h day, whereas subjects with a freerunning period even slightly longer 28 h would become synchronized. In the intraindividual comparison, overt rhythms of different variables show different entrainment limits. For instance, rhythms in urinary excretion of different electrolytes can be dissociated for several days; the same is true with the rhythms of deep body temperature and performance. This temporal separation excludes the possibility of functional interdependencies between the variables under consideration. Consequently, results obtained with this method of fractional desynchronization do not only assist in evaluating properties of the circadian system, but also assist in the search for physiological interconnections between different variables.