Circadian fluctuations in per mRNA and protein are central to the operation of a negative feedback loop that is necessary for setting the free-running period and for entraining the circadian oscillator to light-dark cycles. In this study, per mRNA cycling and locomotor activity rhythms were measured under different light and dark cycling regimes to determine how photoperiods affect the molecular feedback loop and circadian behavior, respectively. These experiments reveal that per mRNA peaks in abundance 4 h after lights-off in photoperiods of <16 h, that phase shifts in per mRNA cycling and behavioral rhythmicity occur rapidly after flies are transferred from one photoperiod to another, and that photoperiods longer than 20 h abolish locomotor activity rhythms and leave per mRNA at a median constitutive level. These results indicate that the per feedback loop uses lights-off as a phase reference point and suggest (along with previous findings for per 01 and tim 01 ) that per mRNA cycling is not regulated via simple negative feedback from the per protein.Circadian rhythms in biochemical, physiological, and behavioral phenomena are a fundamental adaptation of both prokaryotic and eukaryotic organisms to environmental changes that occur over a 24-h period. These rhythms are driven by an endogenous clock that continues to operate under constant environmental conditions. The timekeeping component of the clock, or pacemaker, maintains a periodicity that can be hours longer or shorter than 24 h, and it is synchronized to local time by such environmental signals as light and dark. A stable phase relationship between the pacemaker and its Zeitgeber is clearly a prerequisite if preprogrammed biological changes are to be appropriately timed to daily environmental changes. Physiological and behavioral experiments have been used to determine how the clock adjusts its phase to circadian cycles composed of different proportions of light and dark. During these different photoperiodic conditions, the phase of the circadian pacemaker (and its physiological and behavioral output) is altered so that a stable phase relationship is maintained (26). Because of the lack of measurable pacemaker components, however, these physiological studies could not address the molecular mechanism by which the clock adjusts its phase to accommodate different environmental light-dark (LD) cycles.Genetic screens for rhythm mutants have been used to identify components of the circadian pacemaker. Mutations in the per gene from Drosophila melanogaster can shorten (per S and per) circadian rhythms of locomotor activity and eclosion during constant dark (DD) conditions (16,17) and alter the phase of locomotor activity and eclosion rhythms during LD cycling conditions (3,10,11,17,30). These behavioral effects of the per mutants are paralleled at the molecular level by circadian fluctuations in the abundance of per mRNA and per protein (PER) (12, 41). These fluctuations in per mRNA and protein levels compose a negative feedback loop in which per mRNA serves ...