Background: With the outbreak of the COVID-19 pandemic, pediatric experts called attention to the potential adverse effects of living restrictions (e.g., lockdown) on child well-being, but at the same time– acknowledged their possible benefits. To date, only few data-driven reports have been published on child sleep during COVID-19, and all have been based on parent- or self-reports. This study used auto-videosomnography to capture the effects of COVID- 19 stay-at-home orders imposed in the USA on objectively measured infant sleep. Methods: Auto-videosomnography metrics of infants assessed nightly between January and May 2020 were compared with metrics of an equivalent infant cohort, assessed in the corresponding 2019 period. A total of 610 infants (50.7% girls) aged 6–18 months (M = 11.8, SD = 3.6) were included, with 71,472 analyzed nights. Multilevel models were applied to assess differences between 2019 and 2020 infant sleep pre- and during-lockdown. Results: Whereas infant cohorts were equivalent in demographic and January–March/April sleep characteristics, during the 2020 lockdown infants had longer nighttime sleep durations (Mdifference = 11.0 min, p = .01), later morning rise times (Mdifference = 9.5 min, p = .008), and later out-of-crib times (Mdifference = 12.3 min, p < .0001), compared to the equivalent 2019 period. In addition, weekday-weekend differences in sleep onset and midpoint times were diminished during 2020 home- confinement compared to the equivalent 2019 period (2019: Mdifference = 5.5 min, p < .0001; Mdifference = 4.5 min, p < .0001; 2020: Mdifference = 2.3 min, p = .01; Mdifference = 3.1 min, p < .0001, respectively). Conclusions: Notwithstanding the negative implications of COVID-19 living restrictions in other domains, our findings indicate that there might be a silver lining—in promoting longer and more consistent infant sleep. These benefits should be considered in determining policy for the current and future pandemics. Keywords: Sleep; infancy; coronavirus; home- confinement; videosomnography.