The employees of an oil company in southern England participated in the controlled trial of a saline polyvalent influenza vaccine. Of 981 volunteers, 681 received the vaccine and the remainder an indistinguishable inert control substance. The identity of the latter group was not disclosed until the incidence of influenza had been assessed. Three questionnaires were circulated in December 1963, February and April 1964 for this purpose. Local reactions were significantly more marked in the group which received the vaccine. The pattern of any general reaction did not vary between the groups. Unpleasant reactions would not therefore contra-indicate the use of the vaccine. The year 1963/4 was not an epidemic year for influenza, but the incidence of sporadic cases was similar in the two groups. The severity of the illness, as judged by loss of working time, did not vary between the groups. Inoculation in previous years conferred no additional immunity. This trial suggest that the expense, in terms of time and money, of preventive influenza inoculation is not justified with this preparation. The results might have been different in the presence of an epidemic.