Schistosomiasis, caused by trematodes of genus Schistosoma, is among the most seriously neglected tropical diseases. Although rapid surveillance of risk areas for Schistosoma transmission is vital to control schistosomiasis, the habitat and infection status of this parasite are difficult to assess. Environmental DNA (eDNA) analysis, involving the detection of extra-organismal DNA in water samples, facilitates cost-efficient and sensitive biomonitoring of aquatic environments and is a promising tool to identify Schistosoma habitat and infection risk areas. However, in tropical wetlands, highly turbid water causes filter clogging, thereby decreasing the filtration volume and increasing the risk of false negatives. Therefore, in this study, we aimed to conduct laboratory experiments and field surveys in Lake Victoria, Mbita, to determine the appropriate filter pore size for S. mansoni eDNA collection in terms of particle size and filtration volume. In the laboratory experiment, aquarium water was sequentially filtered using different pore size filters. Targeting >3 µm size fraction was found to be sufficient to capture S. mansoni eDNA particles, regardless of their life cycle stage (egg, miracidia, and cercaria). In the field surveys, GF/D (2.7 µm nominal pore size) filter yielded 2.5-times the filtration volume obtained with a smaller pore size filter and pre-filtration methods under the same time constraints. Moreover, a site-occupancy model was applied to the field detection results to estimate S. mansoni eDNA occurrence and detection probabilities and assess the number of water samples and PCR replicates necessary for efficient eDNA detection. Overall, this study reveals an effective method for S. mansoni eDNA detection in turbid water, facilitating the rapid and sensitive monitoring of its distribution and cost-effective identification of schistosomiasis transmission risk areas.