Phasic pupil responses to melanopsin- and cone-directed stimulation remain stable across circadian time

Human pupil responses vary with time of day, but whether this reflects circadian modulation of specific photoreceptor pathways is unclear. Twelve healthy adults completed repeated pupillometry during a 40-hour ultra-short sleep-wake protocol that limited sustained wakefulness; data from eight met the quality-control criteria and were analysed. This design was intended to limit progressive accumulation of homeostatic sleep pressure across sampled circadian phase. During wake episodes, 0.4-Hz silent-substitution stimuli targeted broadband flux, melanopsin, luminance, L-M cone-opponent, or S-cone pathways. Salivary melatonin provided participant-specific period estimates. We separated slow tonic constriction from phasic pupil tracking and compared hierarchical circadian cosinor models with constant-level models. Constant-level models were favoured for phasic amplitude in all five conditions (log ₁₀ BF ₁₀ = −3.36 to −1.10). They were also favoured for tonic minimum in four conditions; L-M favoured the circadian model (log ₁₀ BF ₁₀ = 0.67), but the evidence was imprecise. Under the tested contrast, adaptation, and frequency, these data do not support circadian modulation of phasic photoreceptor-directed pupil tracking. Other metrics and smaller or non-sinusoidal effects remain possible.