Estimating the range at which harbor porpoises can detect prey items and environmental objects is integral to understanding their biosonar. Understanding the ranges at which they can use echolocation to detect and avoid obstacles is particularly important for strategies to reduce bycatch. Transmission loss (TL) during acoustic propagation is an important determinant of those detection ranges, and it also influences animal detection functions used in passive acoustic monitoring. However, common assumptions regarding TL have rarely been tested. Here, TL of synthetic porpoise clicks was measured in porpoise habitats in Canada and Denmark, and field data were compared with spherical spreading law and ray-trace (Bellhop) model predictions. Both models matched mean observations quite well in most cases, indicating that a spherical spreading law can usually provide an accurate first-order estimate of TL for porpoise sounds in porpoise habitat. However, TL varied significantly (+/-10 dB) between sites and over time in response to variability in seafloor characteristics, sound-speed profiles, and other short-timescale environmental fluctuations. Such variability should be taken into account in estimates of the ranges at which porpoises can communicate acoustically, detect echolocation targets, and be detected via passive acoustic monitoring.