This study examined salmon lice (Lepeophtheirus salmonis) numbers during different full-scale lice treatments of salmon, with focus on detachment during crowding and re-attachment post-treatment. Data was collected from 259 non-medicinal treatments (NMM) during 2018 and 2019. The dataset included data from thermal (Optilicer), mechanical (Flatsetsund Engineering, SkaMik, Hydrolicer) and freshwater treatments. Salmon lice were counted by the staff at the aquaculture sites, and registered in the following categories: sessile, preadult/adult male (PA/AM) and adult female. A counting protocol was established in collaboration with staff from the aquaculture companies taking part, where the time for counting the parasitic salmon lice was: Sample 0: before crowding; Sample 1: at the end of the first crowding; Sample 2: during unloading of the salmon after treatment; and Sample 3: after delousing of the salmon, in the net pen. In addition to the treatments registered directly according to the protocol, we retrieved data for additional treatments directly from the aquaculture companies' own registration system. The objectives of this observational study was (1) investigate whether salmon lice detach during crowding before an NMM; (2) investigate the effect of NMM; and (3) investigate possible re-attachment after delousing. We observed an overall trend of a reduction of salmon lice numbers, with a median percentage decrease of 22% for PA/AM and 29% for adult female lice during crowding. A likely explanation for these observations is that lice detach from the salmon due to mechanical action associated with the increased density of crowding. One-sided Wilcoxon tests corroborated the hypothesis of a reduction in lice counts during crowding for PA/AM and adult female lice (p < 0.05). As expected, we also observed a substantial reduction in lice numbers from Sample 1 to Sample 2, which is most likely due to the effect of the delousing treatment with the NMM. We observed an increase in sessile and PA/AM salmon lice when studying possible re-attachment after delousing. For sessile lice, the increase in lice numbers was most likely due to new infestation. For PA/AM, the increase can primarily be explained by re-infestation with lice that detached during crowding. The increase for PA/AM was greater when the salmon were placed in the same cage (a doubling of lice numbers), compared to placing them in a new cage (p < 0.05), indicating that the salmon lice could re-infest to a greater extent when the salmon were moved a shorter distance. The results presented in this article shed light on the possible dispersal of parasitic salmon lice, which is largely an unexplored topic.