Experimental infection of rainbow trout Oncorhynchus mykiss (Walbaum) with the monogenean Discocotyle sagittata (Leuckart, 1842) allowed comparison between trickle and single exposure, two infection modes demonstrated to occur in the wild. Both types of infection resulted in mean larval attachment success around 50%, which was significantly dependent on dose of infective larvae used (P < 0.0001), but was not affected by mode of infection (P = 0.244). Worms recovered from fish exposed to the same number of oncomiracidia but different mode of infection differed in their rate of development. The developmental stage attained by parasites was significantly affected by number of infective larvae used (P = 0.005), and by the interaction between dose and mode of infection (P = 0.026), suggesting competition among attached larvae. Statistical analysis demonstrated that in the early stages of infestation, worm distribution over the gill arches can be explained by the relative amount of water flowing over them. One, two and three months post-infection parasite numbers were comparable (P = 0.805), but their observed distribution gradually decreased in gill arches III and IV and increased in gill arch I, suggesting that parasites migrate after initial attachment. These results reproduce phenomena observed in the field, indicating that the experimental infection system could be employed to study infection dynamics and host-parasite interactions under controlled conditions.