Almost 6,000 healthy adults inoculated with influenza vaccine were asked to record daily for 5 days the presence or absence of defined local and general symptoms; 70 percent completed and returned the record. The overall incidence of subjective reactions derived in this way was high; approximately 50 percent complained of local pain and 40 percent of general symptoms. Only about one-third of vaccinees recorded no symptoms. The incidence of subjective local reactions was about 8 percent higher in women than men. Among 5,600 vaccinees a total of 63 days sickness absence was attributed to reaction in the week of immunization (1.1 days per 100 employees). Almost all the absences were of less than 3 days duration, unsupported by a doctor's certificate. In the week of vaccination sickness absence among vaccinees was not higher than that of non-vaccinees. Although the incidence of subjective reaction was found to be high and vaccination might itself cause some sickness absence, most of the reactions appear to be trivial. Among 700 employees who were interviewed 1 or 2 days after vaccination in 1973, the incidence of what was judged to be significant local pain and erythema was about 5 percent. Nevertheless,the reactions appear to have an important effect on acceptance of influenza vaccination. From the answers given to a questionnaire 30 percent of employees who refused vaccination did so either because they previously had symptoms after vaccination, or because others had told them of their occurrence. In approximately 16,500 injections, only 2 patients had an acute reaction resembling anaphylaxis.