A series of experiments with Aeromonas salmonicida and infectious hematopoietic necrosis virus (IHNV) were carried out to determine dynamics of the spread of infection in chinook salmon (1.2-1.98g) and rainbow trout (1.2-3.1g). It was found in experiments with A. salmonicida that fish infected by bath immersion became infectious at 4 days postexposure (dpe), one day prior to dying from furunculosis. In cohabitation experiments with a single infected fish donor, an average of 75% disease specific mortality was obtained. There was suggestive evidence that there is a positive relationship between holding volumes and furunculosis prevalence in cohabitation experiments with single donor fish. Median day to infection was inversely correlated with density. The threshold density at density of 1.97 fish/L was approximately 30 times less than the density of 0.08 fish/L, 13.33 and 320 fish respectively. Reproductive ratio (R₀) and transmission coefficient (β) in the furunculosis epizootic were 3.23 and 0.021 (individuals*day)⁻¹ respectively. The mortality rate (α) of infected animals was 28.7% per day. The models constructed successfully mirrored the results of laboratory experiments. Data produced by simulation of the models were significantly associated with the data obtained from laboratory experiments for susceptible (S) class and also for infected class. In similar experiments carried out with IHNV, it was found that donor fish became infectious 3 dpe. The virus levels in donor fish and prevalence levels were also highly associated. Smaller volumes of that led to higher prevalence levels than observed in bigger volumes with 60 or 30 fish in each. A single donor fish was able to transfer virus to recipient fish. However, unlike the A. salmonicida experiment, transmission was insufficient to initiate a full-scale infectious hematopoietic (IHN) epizootic. Estimated parameters for dynamics of infection were approximately half of the values for A. salmonicida (R₀=2.57,β=0.008 (individuals*day)⁻¹ and α=0.15). The models constructed for IHNV spread were used to simulate the results of density experiment. However, it was not possible to test the association between susceptible and infected classes due to inadequate number of infected fish. / Graduation date: 2001