Currently, there are two methods of inducing laboratory-based amoebic gill disease (AGD) in Atlantic salmon, Salmo salar L.: cohabitation with infected ¢sh or exposure to a suspension of amoebae. Amoebic gill disease cannot be induced with cultured amoebae; therefore, the only source of the infective organism is salmon with the disease. For experimental purposes and to maintain pathogen supply, salmon are kept in an infection tank and amoebae are isolated from salmon once the disease establishes. In this way, discrete batches of amoebae are collected periodically. This study investigated the infective ability of di¡erent batches of amoebae. Furthermore, the effect of stocking density of salmon on the progression of AGD was also examined. The infective ability of different batches of amoebae isolated periodically from AGD-a¡ected salmon varied in terms of quanti¢able pathology. Salmon stocking density had a signi¢cant impact on survival after amoebae challenge, with morbidity beginning 23 days post challenge in tanks stocked at 5.0 kg m À 3 and 29 days for those stocked at 1.7 kg m À 3 . For uniform initiation of AGD in multiple tanks, amoebae batches should be equally divided and added to tanks until the required concentration is reached and to maintain a standard biomass between replicate tanks and treatments.Challenge methods: 1, cohabitation of ¢sh with clinical signs of AGD and na|« ve ¢sh; 2, crude water-borne exposure^addition of infective material scrapped from the gills of AGD-a¡ected salmon and partially quanti¢ed; 3, pure water-borne exposure^addition of totally quanti¢ed and partially puri¢ed amoebae isolated from excised gills of AGD-a¡ected salmon. wMorbidity equates to the clinical onset of AGD, i.e. high ventilation rate, loss of equilibrium.Aquaculture Research, 2010, 41, e505^e516 Experimental induction of AGD in Atlantic salmon P B B Crosbie et al.