The photoperiodic system is concealed in the highly complex black-box, comprising four functional subunits: 1) a photo/thermo-sensitive input unit, 2) a photoperiodic clock based on a circadian system, 3) a condenser unit counting the number of inductive signals, and 4) a neuroendocrine switch that triggers a phenotypic shift. This review aims to summarize the research history and current reach of our understanding on this subject to connect it with the molecular mechanism of the circadian clock rapidly being unveiled. The review also focuses on the mode of intersubunit information transduction. It will scan the recent advancement in research on each functional subunit, but special attention will be given to the circadian clock–endocrine conjunct and the role of melatonin signaling in the regulation of insect photoperiodism. Prothoracicotropic hormone (PTTH) probably plays the most crucial role in the regulation of pupal diapause, which is the simplest model system of diapause regulation by hormones investigated so far, particularly in the Chinese oak silkmoth (Antheraea pernyi). A search for the trigger to release the PTTH found some candidates, that is, indoleamines. Indolamine metabolism is controlled by arylalkylamine N-acetyltransferase (aaNAT). Indolamine dynamics and aaNAT enzymatic activity changed according to photoperiods. aaNAT activity and melatonin content in the brain showed not only a photoperiodic response but also a circadian fluctuation. aaNAT had multiple E-boxes, suggesting that it is a clock-controlled gene (ccg), which implies that cycle (cyc, or brain–muscle Arnt-like 1 = Bmal1)/Clock (Clk) heterodimer binds to E-box and stimulates the transcription of aaNAT, which causes the synthesis of melatonin. RNAi against transcription modulators, cyc, or Clk downregulated aaNAT transcription, while RNAi against repressor of cyc/Clk, per upregulated aaNAT transcription. Immunohistochemical localization showed that the circadian neurons carry epitopes of melatonin-producing elements such as aaNAT, the precursor serotonin, HIOMT, and melatonin as well as clock gene products such as cyc-ir, Per-ir, and dbt-ir, while PTTH-producing neurons juxtaposed against the clock neurons showed hMT2-ir in A. pernyi brain. Melatonin probably binds to the putative melatonin receptor (MT) that stimulates Ca2+ influx, which in turn activates PKC. This induces Rab 8 phosphorylation and exocytosis of PTTH, leading to termination of diapause. All the PTTH-expressing neurons have PKC-ir, and Rab8-ir. When diapause is induced and maintained under short days, serotonin binding to 5HTR1B suppresses PTTH release in a yet unknown way. RNAi against this receptor knocked out photoperiodism; short day response is blocked and diapause was terminated even under the short day condition. The result showed that a relatively simple system controls both induction and termination in pupal diapause of A. pernyi: the circadian system regulates the transcription of aaNAT as a binary switch, the enzyme produces a melatonin rhythm that gates PTTH release, and 5HTR1B and MT are probably also under photoperiodic regulation. Finally, we listed the remaining riddles which need to be resolved, to fully understand this highly complex system in future studies.