Termination of breeding, the development of photorefractoriness, subsequent postnuptial molt, and fat deposition precede autumnal migration in White-crowned Sparrows (Zonotrichia leucophrys). The timing of these postreproductive functions differs from that of analogous vernal functions because they are neither synchronous within the population nor uniformly coincident with environmental stimuli. We hypothesize that the expressions of autumnal functions are internally coupled and that they result from a separate process initiated by the increasing vernal daylength. Our data concerning the effect of various photoregimes on testicular regression, postnuptial molt and fat deposition of Z. 1. gambelii are consistent with the following generalizations: (1) on daylengths that exceed 16 h these processes begin after 40-60 days of photostimulation, inversely related to daylength; (2) transfer of photostimulated (20L 4D) males to an intermediate daylength of 12 h near the end of the testicular growth phase blocks the expression of autumnal functions, although their expression is not blocked if the birds are photorefractory when transferred; (3) males transferred from 12L 12D to 20L 4D in various stages of testicular development begin postnuptial molt after a fixed number of days regardless of the stage of testicular development at the time of transfer; (4) rates of postnuptial molt and fat deposition are inversely related to the daylength to which the birds are exposed when these functions occur, which may explain the acceleration of preparation for autumnal migration in late-starting individuals. Therefore, the systems that control autumnal processes appear not only to be independent of those that regulate analogous vernal processes, but also appear to rely more on internal coupling to integrate these functions than do the latter. The physiological preparation that precedes migration usually involves several functions that require coordination--activation (spring) or inactivation (autumn) of the reproductive system, hyperphagia, and fat deposition, and usually at least a partial molt (Stresemann and Stresemann 1966, Farner and Lewis 1971, Murton and Westwood 1977, Wingfield and Farner 1980). These processes occur in a species-specific order and are temporally separated in small avian forms, although the extent of separation varies among species (Stresemann and Stresemann 1966, Payne 1972, Farner et al. 1980). Although the mechanisms that ensure that these functions occur at the proper time and in the proper sequence have been extensively studied in the migratory races of the Whitecrowned Sparrow (Zonotrichia leucophrys), they remain only partially understood. Differences in preparation for vernal and autumnal migration under natural conditions suggest that different control mechanisms are employed in these two seasons. Preparation for spring migration is synchronized within a population of Z. leucophrys wintering at a given latitude because it appears to be initiated in direct response to increasing daylength (Farner and Wilson 1957, King 1961, King and Farner 1965, Mewaldt and King 1978b). In contrast, preparation for autumnal migration is less synchronous within populations because it does not begin until the birds finish nesting. If a clutch or brood is lost, a pair may renest and delay onset of postnuptial molt, although there is a compensatory acceleration of the molt (Wingfield and Farner 1979). In addition, the decreasing daylength of late summer and autumn differs from lengthening days of spring as predictive information for fall migration (King 1963). Because of (1) lack of synchrony among individuals with the same environmental information and (2) individual differences in rate of preparation due to renesting, i seems likely that this sparrow relies on external information to coordinate its post-repro