In studies of the salmon louse Lepeophtheirus salmonis (Krøyer, 1837), experimental design is complicated by a highly variable and unpredictable lice loss among common experimental tanks and a substantial rate of host transfer within tanks. When fish hosting L. salmonis are maintained in individual tanks, unspecific effects such as host transfer, louse predation by cohabitant hosts and agonistic host interactions are excluded. This study suggests that it is possible to maintain Atlantic salmon Salmo salar infected with L. salmonis in an array of small, single fish tanks and, by doing so, provide an experimental system in which the loss of motile pre-adult and adult stages of L. salmonis is predictable. Here, lice can be collected shortly after detachment for detailed studies or to provide mortality curves of lice from individual fish. This represents an experimental approach improving precision in studies of L. salmonis, such as drug and vaccine efficacy assays, RNA interference (RNAi) studies and host-parasite interactions. The natural loss of pre-adult/adult L. salmonis from the system was higher for males than females. The loss of females appeared to be a process somewhat selective against large individuals. Inherent qualities of the host appeared to be of little significance in explaining the variability in loss of preadult/adult lice.