We report an experimental model of autoimmune inflammatory myopathy. Splenic cells from two inbred murine strains (BALB/c and SJL/J) are activated (immunized) in vitro by co-culture with their respective syngeneic skeletal muscle myotubes. Subsequent injection of the activated splenocytes with or without B. pertussis into the respective syngeneic hosts results in inflammatory myopathy in the SJL/J mice but never in the BALB/c mice. The muscle inflammation is very similar in appearance to human autoimmune inflammatory myopathies. The myositis is not effector cell-skeletal muscle specific because splenocytes activated by co-culture with smooth muscle will also elicit skeletal muscle lesions. Both strains of skeletal muscle appear to express class II (Ia) antigens and the splenocytes from both strains appear to be equally activated. Thus we postulate that the difference in the expression of myositis between the two strains is in the effector phase of the disease. Since SJL/J mice have vasoactive amine sensitive vascular systems and BALB/c do not, it is likely that activated splenocytes emigrate from muscle microvessels in the SJL/J strain whereas they cannot do so in the BALB/c strain. The most significant contribution of this model may be in its potential for addressing a sine qua non of cellular autoimmune disease, i.e. lymphocyte migration from the vascular compartment into the target tissue. Finally, the data support a cellular more than a humoral pathogenesis in this model.