The mean infectious doses of selected avian influenza virus (AIV) isolates, determined in domestic poultry under experimental conditions, were shown to be both host-dependent and virus strain-dependent and could be considered one measure of the infectivity and adaptation to a specific host. As such, the mean infectious dose could serve as a quantitative predictor for which strains of AIV, given the right conditions, would be more likely transmitted to and maintained in a given species or subsequently cause an AI outbreak in the given species. The intranasal (IN) mean bird infectious doses (BID50) were determined for 11 high-pathogenicity AIV (HPAIV) isolates of turkey and chicken origin for white leghorn (WL) chickens, and for low-pathogenicity AIV (LPAIV) isolates of chicken (n = 1) and wild mallards (n = 2) for turkeys, and WL and white Plymouth rock (WPR) chickens, domestic ducks and geese, and Japanese quail. The BID50 for HPAIV isolates for WL chickens ranged from 10(1.2) to 10(4.7) mean embryo infectious dose (EID50) (median = 10(2.9)). For chicken-origin HPAIV isolates, the BID50 in WL chickens ranged from 10(1.2) to 10(3.0) EID50 (median = 10(2.6)), whereas for HPAIV isolates of turkey origin, the BID50 in WL chickens was higher, ranging from 10(2.8) to 10(4.7) EID50 (median = 10(3.9)). The BID50 of 10(4.7) was for a turkey-origin HPAIV virus that was not transmitted to chickens on the same farm, suggesting that, under the specific conditions present on that farm, there was insufficient infectivity, adaptation, or exposure to that virus population for sustained chicken transmission. Although the upper BID50 limit for predicting infectivity and sustainable transmissibility for a specific species is unknown, a BID50 < 10(4.7) was suggestive of such transmissibility. For the LPAIVs, there was a trend for domestic ducks and geese and Japanese quail to have the greatest susceptible and for WL chickens to be the most resistant, but turkeys were susceptible to two LPAIV tested when used at moderate challenge doses. This suggests domestic ducks and geese, turkeys, and Japanese quail could serve as bridging species for LPAIVs from wild waterfowl to chickens and other gallinaceous poultry. These data do provide support for the commonly held and intuitive belief that mixing of poultry species during rearing and in outdoor production systems is a major risk factor for interspecies transmission of AIVs and for the emergence of new AIV strains capable of causing AI outbreaks because these situations present a more diverse host population to circumvent the natural host dependency or host range of circulating viruses.