Groups of turkeys were exposed to different isolates of avian influenza virus from wild mallard ducks and domestic turkeys by the intracerebral, intravenous, intratracheal, and intra-airsac routes, and pathogenicity indices were calculated. For the intracerebral pathogenicity study, body weight was also measured. For intravenous, intratracheal, and intra-airsac pathogenicity studies, necropsy lesions were scored and serological responses were recorded. Only the intracerebral pathogenicity index and body weight gain post intracerebral infection demonstrated any differences between isolates. The other procedures failed to demonstrate any pathogenicity whatsoever. There was a correlation (R = 0.73) between intracerebral pathogenicity index and reduced weight gain postinfection. These studies suggest that growth suppression may be an objective measure of pathogenic potential of influenza viruses found to be nonpathogenic by other methods.