Serum specimens from 53 evaluable patients enrolled in a clinical trial of lymphoblastoid interferon in recurrent respiratory papillomatosis were screened for the presence of interferon-binding antibodies by an indirect enzyme immunoassay and evaluated for neutralizing antibody measured as the inhibition of antiviral activity. Immunoglobulin G antibodies that specifically bound lymphoblastoid interferon were detected in 66% (35 of 53) of patients; neutralizing antibody was detected in 11 of the 35 patients having binding antibody (and in none of the patients who were negative for binding antibody). The incidence of detectable neutralizing antibody in this study population was 20.8% (11 of 53), which is markedly higher than in previous reports of lymphoblastoid interferon in patients with other diseases (i.e., less than 1% incidence). The cumulative dose received at the time of detection of neutralizing antibody ranged from 163 to 385 MU per square meter of body surface. Neutralizing antibody was detectable at a median time of 120 days after initiation of interferon therapy, and binding antibody appeared earlier in those patients (median 59 days) than in patients in whom only binding antibody was produced (median 116 days). Despite the tendency of binding antibody to appear either in patients in whom neutralizing antibody was eventually formed, the detection of binding antibody was not necessarily predictive of the subsequent development of neutralizing antibodies. Binding antibody persisted after neutralizing antibodies had become undetectable.(ABSTRACT TRUNCATED AT 250 WORDS)