We have tested the two principal theories which explain the previous finding that small amounts of type-specific antibody to the adenovirus hexon can neutralize infectivity, whereas even large amounts of cross-reactive antibody do not. a) It has been suggested that the type-specific determinants are especially prominent in the virion. We have therefore measured the capacity of whole virus to bind appropriate antibodies, using a sensitive radioimmunoprecipitation (RIP) system. In fact, virions bound type-specific and cross-reactive antibodies impartially. Moreover, they bound both much less effectively than did free hexon or disrupted virus, suggesting that many of each kind of determinant are inaccessible in virions. b) It has been suggested that the type-specific determinants are confined to those hexons located next to the pentons, and that they are the targets for neutralizing antibody. We have therefore studied the antigenicity of peripentonal and nonamer hexons isolated from virions, and found that each possessed both kinds of determinants. Furthermore, these were present in the same proportion as in hexons purified from the soluble antigens in infected cells ("free hexons"). We concluded that the mechanism of neutralization by antibody is complicated, and that the type-specific determinants exposed on the virion must play a crucial role.