During relapses in relapsing-remitting multiple sclerosis (RRMS), serum soluble HLA class I surface antigen (sHLA-I) levels are reported to either decrease or remain unchanged, whereas serum sHLA-II levels increase. Interferon-beta1b therapy was recently reported to increase serum sHLA-I in RRMS. In the present prospective study, solid-phase enzyme-linked immunosorbent assay was used to measure sHLA-I and sHLA-II in the sera of 21 RRMS patients during a clinical exacerbation, and then six weeks after treatment with high-dose interferon-beta1a (IFN-beta1a). Pretreatment serum sHLA-I was significantly lower in patients than in normal controls (P < 0.0005). Pretreatment sHLA-II was also significantly lower than in normal controls (P = .003) unless enhancing MRI lesions (objectified relapse) were present; then sHLA-II levels were similar to normal controls (relative increase). Six weeks after initiation of IFN-beta1a treatment, a significant increase in serum sHLA-I was observed in all 21 RRMS patients (P < .0005). Conversely, serum sHLA-II decreased significantly after treatment in the entire patient group (P < .0005). The acute effect of IFN-beta1a on serum sHLA-I and sHLA-II was observed to be the opposite of that occurring during RRMS relapses. Monitoring of both sHLA-I and sHLA-II appears necessary if these molecules are to be developed as RRMS activity markers.