The results of three serologically controlled double blind field trials in army units are presented. The evaluation of results according to morbidity, regardless of aetiology, showed a significant reduction in acute respiratory diseases (ARD) after administration of Impulsin. In the 1973 trial (901 volunteers), 22.7% of ARD cases were found in the Impulsin group contrary to 34.4% in the placebo group (P less than 0.0002). The relevant values in the 1974 trial (610 volunteers) were 19.7% and 40.7% (P less than 0.002) and in the 1975 trial (353 volunteers) 10.6% and 28.8% (P less than 0.004). The study of the immunological background in representative sets of volunteers allowed determination of the aetiology, the proportion of asymptomatic infections and possible deformation of results due to preexisting protective antibodies. Manifestation rate (MR) expressing the proportion of sick persons out of all sensitive subjects with serologically proved infection was found useful. This indicator is relatively independent of randomization and is more sensitive as compared to the incidence rate. In the 1973 trial, influenza A 2 England was prevalent, the MR of infection being 15.4% in the Impulsin group and 44.9% in the placebo group (P less than 0.0002). After elimination of persons with preinfection antibodies greater than or equal to 1:256 the corresponding values of MR were 17.6% and 46.6% (P less than 0.005), reflecting the "relatively clean effect" of Impulsin. In the 1974 trial, where influenza B Hong-Kong was prevalent, MR was 14.3% and 57.1%, respectively (P less than 0.001). Preinfection antibodies were negligible. The preliminary prophylactic index of the drug seemed to be 4.3 for combined adenoviral infections (trials 1973 and 1974 taken together). In the 1975 trial, the results of serological examination were unsatisfactory. Antibodies vs. influenza A Port Chalmers were found in 24.5% of ARD only. The differnce is aetiologically unclarified ARD was statistically significant. Although displaying a significant limitation of clinical infections, the administration of Impulsin did not seem to have any influence on the formation of antibodies.