A six-wave longitudinal panel survey of a United States (U.S.) online opportunity sample panel occurred between February 28, 2020, with confirmed U.S. SARS-CoV-2 viral infections < 50, and April 13, 2021, a few months after general public vaccinations began. Each survey wave included the same baseline questions—e.g., COVID-19COVID-19 attention, media use, risk perceptionsRisk perceptions, psychological distance, COVID-19COVID-19 knowledge, behavioral intentionsBehavioral intentions, policy supportPolicy support, trust in decision-makers—to control for variation within and across waves, with other variables in specific waves—e.g., demographics, conspiracist thinking, anti-Chinese prejudice, science attitudes. Among other findings, risk perceptionsRisk perceptions, trust, behavior, and policy supportPolicy support varied over time. Results supported behavioral motivation (higher judged risk now means greater protective action later) and accuracy (lower reported protection now means higher risk perception now). However, results strongly refuted risk reappraisal (greater protective action now meant higher, not lower, risk perceptionsRisk perceptions later), perhaps reflecting continuing uncertainties about COVID-19COVID-19 dynamics and protectiveness of policies and behaviors. Risk perceptionsRisk perceptions for the U.S. and the world, and affective risk perceptionsRisk perceptions (e.g., concern, dread) were more strongly associated with behavior and policy supportPolicy support for mask-wearing, avoiding public gatherings, and vaccination than was personal risk perception. These results have implications for future pandemic risk management and communication practice.