Conflicting results have been reported concerning the association between high age and response to influenza vaccines. Some authors have found a reduced response in aged subjects, others have found no difference or even better results as compared with younger control subjects. Seventeen papers were selected from international literature published in the period 1968-1988 for a review of the anti-haemagglutinin-IgG sero-response following vaccination: among 30 cases in which vaccine components could be studied independently, ten revealed a better immune response in young subjects than in the elderly, four found more favourable results in the elderly, and 16 could not detect any significant between-group-differences, the latter most probably because of a high type-2-error. Nine of these 16 cases tended to favour young subjects. These results were relativated by the finding that each paper had at least one of three methodological limitations: (1) the failure to exclude subjects with illnesses or using drugs influencing the immune system, (2) the failure to exclude subjects with previous vaccinations against influenza, (3) the failure to exclude subjects with high prevaccination antibody titres. The direction of these biases is such that failure to address any one issue will lead to an underestimate of the response of aged subjects. In view of the failure to control these biases, it was not surprising that the papers reviewed presented a heterogeneous picture. Thus, the association between high age per se and response to influenza vaccines, if any, has not yet been established. Suggestions are made for future studies in which admission criteria should control health state and previous exposure to influenza antigens.