Cooney, M. K., C. E. Hall and J. P. Fox (Univ. of Washington School of Public Health and Community Medicine, Seattle, Wash. 98195). The Seattle Virus Watch. III. Evaluation of isolation methods and summary of infections detected by virus isolations. Am J Epidemiol 96: 286–305, 1972.—This paper reviews the different cell cultures employed for isolations and describes by major group the frequencies of virus isolations and the incidence of infections detected. Primary human embryonic kidney (HEK) cultures were used for respiratory and fecal specimens whereas WI-38 cell cultures were used chiefly for respiratory specimens. Screening of 19, 000 mainly respiratory specimens in 2 or more types of cell culture showed the cultures' sensitivity for different virus groups. HEK far surpassed WI-38 or combined primary monkey kidney and HeLa cell cultures in recovering polio, coxsackie B, adeno and myxoviruses but was insensitive for rhinoviruses, and poorly sensitive for respiratory syncytial (RS) virus. WI-38 cultures were far superior to HEK for detecting rhinoviruses, herpesviruses and, surprisingly, echoviruses. Special cultures were necessary to recover RS virus and cytomegalovirus. Judged by overall isolation rates, the total screening system was more efficient than that used in the New York Virus Watch (VW); its sensitivity is also shown by comparing the frequency with which 2 or more nonpolioviruses were recovered from the same specimen with that expected based on overall isolation rates. Discovery of 2 or 3 viruses in 31 specimens required 2 types of cell culture. Examination of 18, 000 respiratory and 14, 500 fecal specimens yielded 1, 425 poliovirus isolates and 1, 426 isolates from 922 nonpoliovirus infections. Poliovirus isolation rates (presumably of vaccine origin) were 0.8% from respiratory and 9.6% from fecal specimens from index persons, and under age 2 were 1.4% and 13.8%, respectively. Isolation and incidence rates were inversely related to age except for herpesviruses. The most frequently recovered nonpolioviruses from respiratory specimens were rhino and adenoviruses (2.1% and 0.9%, respectively, overall and 3.1% and 1.5% under age 2). Adenoviruses were recovered 3 times more frequently from fecal specimens (2.4%) than either coxsackie or echoviruses. On a person-year basis, incidence rates for all index persons and for those under age 2, respectively, were: all nonpolioviruses, 1.56 and 3.1; rhinoviruses, .59 and 1.21; adenoviruses, .33 and .88. Concurrent shedding of 2 or more nonpolioviruses in pairs of respiratory and fecal specimens were noted in 76 instances, most often chronically excreted (adeno and herpes) or frequently recovered (rhino) viruses. Isolation rates examined in relation to interval between specimen date and illness onset support the presumption, previously stated on the basis of the New York VW data, that isolates marking the initiation of an infection are causally related to illnesses with onsets within 7 or (for rhinoviruses) 2 days after to 13 days before the positive specimen date. Such presumption is necessary in estimating the frequency with which infection with specified agents results in disease.