Uncontrolled forest fire is an important driver of forest degradation throughout the country. Recurrent forest fires severely damage and prohibit regeneration and growth of seedlings, destroy non-timber forest products and, in some cases, encourage invasive species. Although quantitative information is not available, forest fires are definitely degrading biodiversity, enhancing soil erosion and inducing floods and landslides due to the destruction of the natural vegetation in Nepal. In this study, we scrutinized forest fires activity in Nepal using the Moderate Resolution Imaging Spectroradiometer (MODIS) active fire datasets. We assessed fire magnitude, distribution, seasonality and risk zone in diverse geographical area. MODIS data from 2000 to 2013 revealed an average 2159 fire counts per year with the highest during 2012. Fire season in Nepal extends from January to June with the peak during April. Using Kernel density model, identified hotspot regions of fire in diverse regions of Nepal. Multi-collinearity diagnostic procedure was implemented by examining tolerance and Variance Inflation Factor (VIF). VIF values for each of the predictor were below two, which suggests low or moderate multi-collinearity. These results on forest fires will be useful to stakeholders to address the impact of forest fire in Nepal.