The novel coronavirus disease 2019 (COVID-19) has been the biggest public health problem of the present world. As the number of people suffering from the pandemic is rising, it is likely to claim more life and worsen the global health and economy. Nepal, one of the developing countries in the south Asia has been strongly influenced by the pandemic and struggling to contain it with multiple interventions, however, spatiotemporal dynamics of the epidemic and its linkage with various intervention strategies has not been studied yet. Here, we employed the prospective spatial-temporal analysis with SaTScan assessing dynamics of the COVID-19 cases from 23 January to 31 August 2020 at district level in Nepal. The results revealed that COVID-19 dynamics in the early stage of transmission was slower and confined in certain districts. However, from the third week of April, transmission spread rapidly across districts of Province No. 2 and Sudoorpaschim Province, primarily introduced by Nepalese citizens returning from India. Despite nationwide lockdown, nine statistically significant active and emerging clusters were detected between 23 January and 21 July 2020, whereas ten emerging clusters were observed for extended period to 31 August. The population density and population inflow from India crossing the sealed border had significant effects on the elevated risk of the epidemic. The capital city Kathmandu has become the highest-risk active cluster since August when travel restriction has been suspended. Movement restriction appears to be the most effective non-pharmaceutical intervention against the COVID-19 for resource-scarce countries with limited health care facilities. Our findings could be valuable to the health authorities within Nepal and beyond to better allocate resources and improve interventions on the pandemic for containing it efficiently.