Cell populations of Paramecium bursaria show mating reactivity in the light period, but not in the dark period, when exposed to a light-dark cycle (LD 12:12). After they are transferred to constant-light (LL) conditions (1,000 lux), they continue to show a circadian rhythm of mating reactivity. The rhythm gradually dampens in LL so that mating reactivity in populations becomes arrhythmic in LL within 2 weeks. We wanted to know whether the arrhythmicity of this population was due to the absence of circadian rhythmicity within each individual cell, or merely due to asynchrony of a population of individually rhythmic cells. Therefore, single cells were isolated randomly from an arrhythmic population that had been in LL for a long time. Then the mating reactivity of these single cells was individually tested every 3 hr for 2 days. Each single cell showed a circadian mating rhythm in LL. This shows that the disappearance of the mating rhythm in cell populations under LL is not caused by disappearance of circadian rhythmicity within individual cells, but is due to desynchronization among cells in a population. When an arrhythmic population in LL is darkened for 9 hr, the mating reactivity rhythm of the cell population reappears. This occurs by resynchronization of the rhythms among individual cells, as can be shown by exposing single cells to pulses of 9 hr of darkness. This dark treatment causes phase shifts of single-cell rhythms, and a phase response curve is obtained for this stimulus. This phase-shifting behavior explains the efficacy of 9-hr dark pulses in restoring the population's rhythm.