Estimating the population size of shy and elusive species is challenging but necessary to inform appropriate conservation actions for threatened or declining species. Using camera-trap surveys conducted during 2017–2021, we estimated and compared African clawless otter Aonyx capensis population densities and activity times in six conserved areas in southern Africa. We used two different models to estimate densities: random encounter models and camera-trap distance sampling. Our results highlight a general pattern of higher estimated densities and narrower confidence intervals using random encounter models compared to camera-trap distance sampling. We found substantial variation in densities between study areas, with random encounter model estimates ranging between 0.9 and 4.2 otters/km2. Our camera-trap distance sampling estimates supported the relative density estimates obtained from random encounter models but were generally lower and more variable, ranging from 0.8 to 4.0 otters/km2. We found significant differences in otter activity patterns, with populations either being nocturnal, mostly nocturnal or cathemeral. As all study areas experience little human disturbance, our results suggest that there are large natural variations in otter densities and activity patterns between regions. When densities are converted to metrics that are comparable to previous studies, our estimates suggest that African clawless otter population numbers are generally lower than previously reported. This highlights a need for broader spatial coverage of otter population assessments and future studies to assess potential environmental drivers of spatial, and potentially temporal, variation in population numbers and activity patterns.