Using changes in type-specific complement-fixation (CF) antibody and, to a lesser extent, isolation of virus, the Seattle Virus Watch (VW) monitored the activity of influenzaviruses in the community for nearly four years (1965–1969). Prevalence of CF antibody to both A and B varied greatly and in nearly parallel fashion, closely reflecting recent infection experience but providing no basis for correct prediction of outbreaks, e.g. despite 96% prevalence in summer 1969, a second major Hong Kong (HK) outbreak occurred in winter 1969–1970. Concurrent or closely consecutive waves of infection with viruses of both types occurred in the winters of 1965–1966, 1967–1968 (community recognition focused on A2) and 1968–1969 (B outbreak exceeded HK in size and duration). However, well documented infections with both A and B occurred “off season” (April through November) in three years, as well as in winter 1966–1967. Somewhat more B than A infections were detected: 221 vs. 199, giving estimated rates per 100 person years of 19.9 and 18.6. Highest age-specific infection rates for both viruses were in the 2–5-year age group (26.5 for A, 28.0 for B) with next highest in infants (24.4 and 26.3, respectively). Much lower rates were observed in older children (18.0 and 15.0) and adults (14.1 and 16.8). Within invaded families the age patterns of infection for A2, HK and B were similar with the highest rates under age 6 and much lower rates in older children, suggesting community spread via preschool rather than school children. Based on estimated secondary attack rates, B was more infectious (31.6%) than either A2 (28.3%) or, surprisingly, the new HK variant (24.1%). This difference in apparent infectivity is explained largely by the much greater frequency with which reinfection with the same major virus subtype occurs with B than with A (48% of B infections were reinfections vs. a maximum of 36% of A2), a phenomenon which helps explain how influenza B has persisted without undergoing the periodic major antigenic changes characteristic of influenza A. Reinforcing the signficance of influenza B reinfections is their demonstrated association with virus shedding and illness. Because of insensitive virus isolation methods, the significance of reinfections wth A2 could not be similarly evaluated.