Evaluating the influence of observed daily weather on observed fire-related effects (e.g. smoke production, carbon emissions and burn severity) often involves knowing exactly what day any given area has burned. As such, several studies have used fire progression maps - in which the perimeter of an actively burning fire is mapped at a fairly high temporal resolution - or MODIS satellite data to determine the day-of-burning, thereby allowing an evaluation of the influence of daily weather. However, fire progression maps have many caveats, the most substantial being that they are rarely mapped on a daily basis and may not be available in remote locations. Although MODIS fire detection data provide an alternative due to its global coverage and high temporal resolution, its coarse spatial resolution (1 km2) often requires that it be downscaled. An objective evaluation of how to best downscale, or interpolate, MODIS fire detection data is necessary. I evaluated 10 spatial interpolation techniques on 21 fires by comparing the day-of-burning as estimated with spatial interpolation of MODIS fire detection data to the day-of-burning that was recorded in fire progression maps. The day-of-burning maps generated with the best performing interpolation technique showed reasonably high quantitative and qualitative agreement with fire progression maps. Consequently, the methods described in this paper provide a viable option for producing day-of-burning data where fire progression maps are of poor quality or unavailable.