Individuals with neuromuscular disease show a wide spectrum of muscle pathology. To test the hypothesis that the immune response to trivalent inactivated influenza vaccine is potentially inadequate when given intramuscularly into a fibrosed muscle, this prospective randomized study compared the immunogenicity and safety of the standard intramuscular versus subcutaneous administration of the influenza vaccine in 22 nonambulatory subjects, of whom 10 have been on glucocorticoid therapy. Analysis of hemagglutination inhibition antibody titers showed high prevalence of seroprotection (prevaccination of 82% H1N1, 72% H3N2, 31% B; postvaccination of 100% H1N1, 77% H3N2, 59% B). Geometric mean titer ratios for each antigen showed no significant difference (P > .5) between intramuscular and subcutaneous routes. Seroprotection was not adversely affected by glucocorticoid therapy. Local tolerance was better with subcutaneous route. Larger studies are needed to confirm these preliminary results.