The complete nucleotide sequences of full-length copies of genomic segment 6 or M3 of US bluetongue virus serotype 11 and 17 consisted of 1638 nucleotides. The plus-strand contained an open reading frame for a protein of 526 amino acids which was equivalent to about 59,000 Da, similar to the molecular weight of GP5 as determined by SDS-PAGE analysis. This long open reading frame was flanked by a 5' non-coding region of 29 nucleotides and a 3' non-coding region of 28 bases. When the predicted amino acid sequences of GP5 of BTV-11 and -17 were aligned and compared with those of BTV-2, -10, -13, -1AU and -1SA, four major highly conserved domains interrupted by several variable regions were detected. The potential significance of these discrete domains is discussed. Evolutionary and phylogenetic characteristics of these US BTV serotypes were consistent with our finding concerning BTV-1AU and -1SA.