We have previously reported the nucleotide sequences of gene 2 (spike (S) protein gene), gene 6 (nucleocapsid (N) protein gene), and the 3′ end untranslated region of a novel avian infectious bronchitis virus (IBV) strain, CU-T2 [Jia et al. (1995) Arch. Virol. 140, 259–271]. In the present report we describe the sequences of the remaining genes of this strain (gene 3, 4 and 5) with the exception of gene 1 (RNA polymerase gene). Gene 3 contained three open reading frames (ORFs), 3a, 3b and 3c of 174, 195 and 282 nucleotides (nt), respectively. Gene 4 (membrane (M) protein gene) consisted of 749 nt with a single ORF of 687 nt. Gene 5 contained two ORFs, 5a and 5b, with 198 and 249 nt, respectively. Thus, in total, there were 7349 nt from the 5′ end of S protein gene to the 3′ end of the CU-T2 genome. The overall nt sequence homologies between gene 3, 4, and 5 of CU-T2 and those of other strains were between 84.1–90.8%, 85.8–88.8% and 90.4–96.4%, respectively. The predicted amino acid (aa) sequence homologies revealed that gene 3b and 5b were more conserved than 3a, 3c and 5a. Each individual gene of CU-T2 strain (with the exception of the RNA polymerase gene) had a different level of homology with the homologous gene of other strains, suggesting that the evolution of IBV strains in general has been a complex, and as yet, poorly understood process. © 1997 Elsevier Science B.V. All rights reserved.