With the aid of published information, we have developed a standardized and reproducible experimental model of furunculosis in rainbow trout (Salmo gairdneri). The infective doses (LD50 = 200 to 2000 germs, i.m.) employed strains of Aeromonas salmonicida, the virulence of which was maintained by passage in 15-g fingerlings. The number of viable cells in the dose was conveniently determined using the drop-plate enumeration technique; however, meaningful cell counts could only be obtained if broth cultures used for infecting the fish were harvested early in the exponential growth phase (OD < 1.000 at 525 nm). Better results were obtained with intramuscular injection than with intraperitoneal injection. The infection procedure involved injecting a dose of 10 LD50, intramuscularly, into each of 30 fish held at 15 °C and recording the mortalities for 10 d. Protection tests in which Tribrissen (28 mg sulfadiazine and 5.6 mg trimethoprim/kg fish for 8 d) was fed or tetracycline (1 mg per fish) i.m. injected into the infected fish served to demonstrate the value of the model. Results were in agreement with field observations with no death for treated fish and a mortality of 96%, for untreated fish. Key wordss rainbow trout, Salmo gairdneri; Aeromonas salmonicida, furunculosis, experimental infection