Two major groups of respiratory syncytial virus (RSV) strains, A and B, have been identified and their patterns of isolation determined in different communities but not simultaneously in multiple communities. In this study, we tested 483 RSV isolates from 14 university laboratories in the United States and Canada for the 1984/1985 and 1985/1986 RSV seasons; 303 (63%) isolates were group A, 114 (24%) were group B, and 66 (14%) could not be grouped. Isolates were subdivided into six subgroups within group A and three within group B; up to six and often four or more different subgroups were isolated in the same laboratory during the same RSV season. The pattern of group and subgroup isolations varied among laboratories during the same year and between years for the same laboratory. These differences suggest that RSV outbreaks are community, possibly regional, but not national phenomena. The ability to identify group and subgroup differences in isolates is a powerful tool for epidemiologic studies of RSV.