Assessing individuals' knowledge and preventive behaviors towards the Coronavirus Disease of 2019 (COVID-19) is essential for the related public health surveillance strategies. Although some of the studies were conducted in Bangladesh, none of these studies considered the geographical distribution of knowledge and preventive behaviors towards COVID-19. Therefore, the present nationwide cross-sectional study with 10,067 samples for the first-time aims to assess the knowledge gap by presenting the geographical distribution of the COVID-19 knowledge and preventive behaviors across all administrative districts of Bangladesh. The measures included socio-demographics and questions about knowledge and preventive behaviors related to COVID-19. One-way ANOVA, independent t-test, and multiple linear regression were used to analyze the data. In addition, GIS-based mapping identified district-wise distribution of the outcomes. Results indicated that the overall mean score of knowledge related to COVID-19 was 14.363 ± 3.073, whereas 16.95 ± 2.89 was for preventive behaviors. Participants' being male, being divorced or widowed, consuming alcohol, smoking cigarettes, living in villages, and having no formal education reported lower performing preventive COVID-19 behaviors. Those participants with higher knowledge scores reported higher preventive COVID-19 behaviors (β = 0.053, p<0.001). However, the model predicted only 13.2% of the variation in preventive COVID-19 behaviors while the overall model being significant. The findings suggest that the Bangladeshi government should initiate appropriate far-reaching program of health education focusing on knowledge and preventive behaviors towards COVID-19 at a community level. After all, the strategies to combat COVID-19 will require individuals' involvement to control and prevent the disease outbreak, for which education is essential.