The developmental transition of the meristem from vegetative to reproductive growth is controlled by the cyclic alternation of light and darkness in photoperiodic plants. Photoperiod is perceived in the leaves or cotyledons, where a flower-inducing signal is produced and transmitted to the apex. To begin to understand the molecular basis of the photoperiodic induction of flowering, we investigated changes in gene expression at the level of mRNA abundance that occur in association with dark induction of flowering in the short-day species Pharbitis nil. Several cDNAs were isolated that corresponded to mRNAs whose abundance is altered after the transition to darkness. The pattern of increase in mRNA levels corresponding to one cDNA clone, PN1, showed a dark-induced maximum at 8 h of darkness, whereas a second clone, PN9, showed a dark-induced accumulation of mRNA with peak levels at 12 to 16 h of darkness. When plants were held in continuous darkness, both PN1 and PN9 exhibited rhythmic patterns of mRNA accumulation with an approximate circadian periodicity, suggesting that their expression is under the control of an endogenous clock. The observed pattern of expression of PN1 and PN9 in cotyledon tissue was unusual in that darkness rather than light promoted mRNA accumulation, which is a temporal pattern of expression distinct from that of several other Pharbitis genes, including Cab, PsaG, and actin, whose mRNAs were most prevalent or equally prevalent in the light. Brief illumination of an inductive dark period by a red light night break strongly inhibited the accumulation of both PN1 and PN9 mRNA. The expression of both PN1 and PN9 was spatially regulated in that mRNA transcripts were detected in the cotyledons and stems, but not the roots, of photoperiodically competent seedlings. Both PN1 and PN9 appeared to be present as single-copy genes in the Pharbitis genome. Sequence analysis has not determined the identity of these genes. Overall, the accumulation of mRNAs corresponding to both PN1 and PN9 closely paralleled the process of photoperiodic floral induction in P. nil, but a clear involvement with this process cannot be established from our findings because of the difficulty of separating photoperiodic events from other light-regulated processes, especially those involved in photosynthesis, such as Cab gene expression. These results identify the products of circadian-regulated genes in photoreceptive tissue of P. nil and support the concept that circadian-regulated gene expression interacting with darkness may be involved in the regulation of photoperiodically controlled physiological processes, including flower induction.