The purpose of this study was to evaluate the spatiotemporal evolution of ungulate-vehicle collision (UVC) hotspots in response to major road construction. Our approach was novel in that we tested the paired use of a clustering method known as Kernel Density Estimation Plus (KDE+) and a spatiotemporal stepwise modification of this method, STKDE+, to monitor UVC hotspots. Two different locations and scales in the province of Alberta, Canada were examined: (1) Deerfoot Trail, a major artery into the City of Calgary which included wildlife mitigation (4.5 kms of fencing with underpasses); and (2) 54.5 km of Highway 63 in northern Alberta which was converted from a 2-lane undivided to a 4-land divided highway. Using government police collision and carcass data (2000–2021), a spatiotemporal analysis was applied using a 3-year moving window with a one-day step to assess hotspots. Before-after or control-impact analyses were used to assess changes to ungulate-vehicle collision frequency (UVC/km/year). On Deerfoot Trail STKDE+ detected emerging UVC hotspots, including a stable hotspot at the south end of the wildlife exclusion fence and ephemeral hotspots within the exclusion fence zone. Wildlife mitigation measures along Deerfoot Trail had a large effect on reducing ungulate-vehicle collisions. At a larger scale, however, net benefits were impacted by UVCs on a nearby unmitigated highway. As the average annual daily traffic (AADT) on Deerfoot increased, UVC shifted temporally to days with lower traffic volume (weekends in this case). Before-after analysis of the conversion of Highway 63 to a 4-land divided highway showed slight, but not significant increases in ungulate-vehicle collision frequency and severity. The majority of historical UVC hotspots re-emerged after construction was completed. Our analysis highlighted the need to incorporate wildlife considerations at a variety of scales into the transportation planning process. We demonstrated the complimentary nature of different UVC clustering approaches as part of retrospective UVC analyses or for dashboard-based real-time monitoring of UVC hotspots.