Long-term monitoring of species is necessary for estimating trends in abundance and distribution, andassessing impacts of human activities and conservation actions. As the monitoring needs often exceedavailable resources, optimisation of monitoring effort is important. As part of the Swedish nationalmonitoring programme of harbour porpoises, echolocation activity has been measured at 14 stationslocated in five Natura 2000 sites in the Kattegat Sea over four years (2019–2023). We aimed to evaluatewhether monitoring effort could be reduced while still providing the power to detect a 30% decline overthree generations (equivalent to an annual decline of 1.6%) with 80% power and 20% significance, i.e.similar to OSPAR’s threshold for abundance of cetacean assessment units. Stations were divided intofour groups (3–5 stations/group) based on their spatial distribution and the level and variation in the ratioof the weekly buzz-positive minutes (BPM) to detection-positive minutes (DPM). The BPM metrics wereassumed to indicate differences in habitat use. For each station, daily DPM was modelled as a functionof the Julian day and year to create annual indices. For each group, the trend of the annual indices andassociated power was calculated using a log-linear model over the four years. Finally, the power todetect a trend over time using 1–3 stations per group was estimated, assuming three levels of variance.The variance levels chosen were similar to the lowest, mean and highest variance observed for the 14stations over the four monitoring years. For all four groups of stations, a significant trend was estimatedto be detected within 12 years. The power analysis showed that decreasing the number of stations withina group from 3 to 2 delays the time to reach 80% power by approximately 1 year, while decreasing it from2 to 1 delays it by approximately 2 years. Alternating stations within a group has negligible impact on thepower. Depending on the assumed level of variance, a significant trend could be detected using 2 stationsper group within 6 (σ=0.05), 12 (σ=0.15), or 18 years (σ=0.30). The results indicate that monitoring effortcan be moved to other areas, while still providing sufficient power to allow declines in detections to bedetected within 12 years. Further, they demonstrate that clear objectives and evaluation of the effortneeded are crucial for optimisation among areas, populations or species when resources are limited.