In early 2020, governments around the world imposed various non-pharmaceutical interventions (NPIs) to slow the spread of COVID-19. NPIs like stay-at-home (or “lockdown”) orders were especially effective in mitigating virus transmission but could also significantly disrupt people's usual activities. Given the hypothesized relationship between routine disruption and stress (e.g., Hou, Lai, Ben-Ezra, & Goodwin, 2020), we predicted that the amount of daily-life disruption people experienced during lockdowns would significantly predict their stress levels. To test this hypothesis, we collected data from over 300 adults living across the United States via an online survey. Analyses revealed that lockdowns did disrupt participants’ daily routines, that most participants experienced several stress symptoms during lockdowns, and that overall life disruption strongly correlated with stress scores (r = .50). Subsequent multiple linear regression analyses revealed that changes in just 4-5 activities were driving this relationship—exercising, relaxing, using social media, eating, and volunteering—and that disruptions to these behavioral routines predicted between 24-32% of the variance in participants’ stress symptoms. Overall, our results showed that many disruptions in daily routines correlated with stress during lockdowns but that some activity changes were more strongly related to stress than others.