An acute exacerbation of bronchitis is preceded sometimes by a common cold or influenza-like illness, and for this reason viruses have been incriminated as the initiating factor in such attacks. During the past 10 years many workers have attempted to elucidate the role of viruses in exacerbations of chronic bronchitis, but the results have been contradictory and attempts to isolate a virus have been particularly disappointing. In a survey carried out by Murdoch, Leckie, Downie, Swain, and Gould (1959) on outpatients in Edinburgh, 14 out of 31 patients showed a rise in complement-fixing antibodies to either influ enza A, B, or C or adenovirus, but only three of these rises could be associated with clinical infection. Jack and Gandevia (1960) reported completely negative virological findings in 19 patients during one year. Stark, Heath, and Curwen (1965) examined sera from outpatients in London at four-week inter vals: 23 out of 199 patients showed evidence of viral infection, and in 13 this could be correlated with an exacerbation. A more comprehensive study was carried out by Carilli, Gohd, and Gordon (1964) in Boston: 46 cases of respiratory illness occurred, and 23 of these could be associated with viral infec tions ; respiratory syncytial virus alone appeared to be respon sible for 17%, but no common-cold viruses were isolated. Sommerville (1963) has also shown a high incidence of anti bodies to respiratory syncytial virus in sera from chronic bronchitics in Glasgow. More recent studies in the Glasgow area showed 16% of exacerbations to be associated with rising titres to various respiratory viruses (Ross, McMichael, Eadie, Lees, Murray, and Pinkerton, 1967). The patients included in the present survey had been included in the Medical Research Council Early Bronchitis Trial, and the clinical and bacteriological findings will be published separately (Calder and Schonell, 1967). Sera have been taken from these men at approximately eight-week intervals since December 1961 and have been tested for complement-fixing antibodies to the respiratory viruses to detect evidence of infection. The survey also provided the opportunity to study the length of time for which antibody titres persist in patients with chronic bronchitis.