ABSTRACTThe dynamic characterization of the COVID-19 outbreak is critical to implement effective actions for its control and eradication but the information available at a global scale is not sufficiently reliable to be used directly. Here, we integrate multiple data sources through dynamical constraints to quantify its temporal evolution and controllability around the world and within the United States. Overall, the numbers of actively infectious individuals have remained high beyond targeted controllability, with worldwide estimates of 10.24 million on November 24, 2020, totaling in 266.1 million cumulative infections growing at a rate of 11.12 million new infections per week. The actively infectious population reached a local maximum of 7.33 million on July 16, 2020 and remained virtually stagnant at a global scale, with growth rates for most countries around zero that compensated each other, until reverting to net growth on September 22, 2020. We validated the approach, contrasting with prevalence data and the effects of nonpharmaceutical interventions, and we identified general patterns of recession, stabilization, and resurgence. The diversity of dynamic behaviors of the outbreak across countries is paralleled by those of states and territories in the United States, converging to remarkably similar global states in both cases. Our results offer precise insights into the dynamics of the outbreak and an efficient avenue for the estimation of the prevalence rates over time.