The World Health Organization declared COVID-19 pandemic outbreak a global concern, due to the rapid spread of the disease and the United Nations described it as a social, human and economic crisis. In recent years, studies about COVID-19 pointed specific risk factors as agents that increase the risk of disease or infection. Hence, the present study aims to identify and determine the influence of potential risk factors contributing to the COVID-19 incidence rates, considering different spatiotemporal variation in Mozambique employing spatial regression models. Data on demographic, educational, environmental and socioeconomic factors were collected and analyzed to explain spatial and temporal variability COVID-19 incidence risk. The developed models are grouped in global (Ordinary Least Squares, Spatial Lag Model and Spatial Error Model) and local regression models (Geographic Weighted Regression and Multi-scale Geographic Weighted Regression). At province level, the following variables were identified as risk factors: Literacy level and Health access points rates in the first wave; Malaria, Education, Literacy and HIV rates in second wave; and HIV and doctors per health access points units rates in third wave, with R2 greater than 79\%. Interestingly, literacy is a protective factor that might be directly related to health literacy and compliance with public health measures. At district level, the models pointed out Education Access Points rate with R2 values of 78.1\%, 74.6\% and 78.4\% in first, second and third waves respectively, as risk factor. Our results demonstrate that literacy levels, health access units, education settlements and incidence of malaria and HIV significantly explain the spatial variability of COVID-19 incidence rates providing useful insights to policymakers for targeted interventions.