Recent evidence points toward an association between higher non‐visual sensitivity to light and a later circadian phase in young adults complaining of a delayed sleep schedule. Light exposure in the evening may therefore induce a larger suppression of melatonin production in these individuals, which might: (a) bias home estimates of melatonin onset; and (b) decrease sleep propensity at bedtime. In this study, we compared home and laboratory melatonin onsets and production in sleep‐delayed and control participants, using saliva samples collected in the 3 hr preceding habitual bedtime. The mean light intensity measured during saliva sampling at home was ~10 lux in both groups. Melatonin suppression at home was significant, averaging 31% and 24% in sleep‐delayed and control individuals, respectively. Group difference in melatonin suppression was not significant. Estimates of melatonin onset were on average 27 min later at home than in laboratory conditions, with no group difference. Looking specifically at sleep‐delayed participants, there was no correlation between non‐visual sensitivity to light and home–laboratory differences in melatonin onsets. However, higher light sensitivity was associated with greater melatonin suppression in the hour before habitual bedtime. Greater melatonin suppression before bedtime was also associated with a later circadian phase. These results indicate that the validity of home estimates of melatonin onset is similar in sleep‐delayed and in control individuals. Results also suggest that increased non‐visual sensitivity to light could impact melatonin secretion in sleep‐delayed individuals and contribute to a late bedtime by delaying circadian phase and decreasing sleep propensity.