Non-neutralizing factor VIII (FVIII) antibodies (FVIII-Ab) in hemophilia A may be associated with an abnormal clinical response to FVIII concentrates. Patients with FVIII inhibitors may develop noncoagulation FVIII-Ab after the induction of immunotolerance. Natural FVIII-Ab may be detected in the plasma of some healthy subjects. The aim of this study was to analyze the presence of FVIII-Ab in the plasma of 53 normal blood donors and 124 patients with hemophilia A (18 patients had a previous history of FVIII inhibitor, but only 12 had inhibitor at the moment this study was performed). FVIIII inhibitor was measured using the Bethesda method. FVIII-Ab were analyzed by a specific ELISA assay using purified FVIII from a monoclonal concentrate and a standard plasma containing 26 Bethesda units (BU) of FVIII inhibitor. Purified FVIII was used to coat wells of a microtiter plate and was incubated with dilutions of plasma to be tested. Bound human IgG FVIII-Ab were detected by incubation with polyclonal sheep anti.human IgG alkaline phosphatase conjugate, and the OD405 was quantitated. A linear fit was obtained (by plotting FVIII-Ab positivity [OD 405nm] versus BU titer) when serial dilutions of this standard inhibitor plasma, containing titers of 0.5 BU or higher, were used. Four different levels of FVIII-Ab positivity [OD 405nm] were distinguished in this assay: Negative levels (-) were obtained with dilutions of the standard inhibitor containing < 0.5 BU. Mild levels (+) were obtained with dilutions of 0.5-5 BU. Moderate levels (+2) were obtained for dilutions ranging from 5-25 BU. Maximum positivity (+3) was obtained for dilutions of titers > 25 BU. FVIII-Ab positivity was detected in eight of the normal subjects (15%): three were found to be moderately positive (+2) and five mildly positive (+). No inhibitory activity was detectable when whole plasma was used. All the hemophilic patients with a presence of FVIII inhibitor at the time of the study were found to be positive for FVIII-Ab. In addition, the level of positivity correlated with the corresponding BU. Four of the six patients who had a history of inhibitory were negative and two positive. Twenty additional patients (16.12%) in whom no inhibitory activity was detected were found to be positive for FVIII-Ab: 16 + and four +2. The mean age of patients with FVII-Ab positivity was significantly higher than that of patients of the FVIII-Ab negative group (p < 0.005). In conclusion, FVIII-Ab positivity in patients with hemophilia A was 17.7% higher than the level of positivity detected by an inhibitory assay. We propose that this method for FVIII-Ab analysis could be used for patients with hemophilia A, at least to complement the functional inhibitor assay. FVIII recovery or half-life should be assessed in patients who test positive for FVIII-Ab and who show no evidence of inhibitor.