The gene encoding the major capsid protein (hexon) of human adenovirus type 40 (Ad40) has been isolated and sequenced. Comparison of the predicted amino acid sequence of the Ad40 hexon with the corresponding polypeptide of the human enteric adenovirus, Ad41, reveals an overall identity of 88%. The majority of the changes in sequence are located in two areas, amino acids 131 to 287 and 390 to 425. Regions in the hexon protein that vary between Ad40 and Ad41 (subgroup F) were the same regions that varied between Ad2 and Ad5 (subgroup C) suggesting that these areas of the protein represent type-specific antigenic determinants. Other areas were conserved within members of a subgroup but varied between subgroups. Fitting of the Ad40 hexon sequence to the known three-dimensional structure of the Ad2 hexon demonstrates that the variable regions are located in the 1(1), 1(2) and 1(4) loops that form the surface of the virion. Of major significance is the absence in Ad40 of the highly acidic region present in both Ad2 and Ad5. In Ad2 this region stretches down into the D-strand of the beta-barrel forming the P1 domain. Molecular modelling indicates that the amino acids in Ad40 which correspond to the acidic region of Ad2 can also be accommodated in the eight-stranded beta-barrel, thereby maintaining the integrity of the barrel. Since the acidic region is also absent from the hexon of Ad41, the sequence of amino acids that replaces the acidic residues may be responsible for some of the distinctive biological properties of the subgroup F adenoviruses.