To clarify the response of frozen shoulder (FS) to intraarticular injections of high-molecular-weight sodium hyaluronate (HA), a mixture of 2.5 ml of HA and 1.5 ml of 1% lidocaine was injected into the glenohumeral joint of 11 patients with FS, 8 of whom received five weekly injections. The patients were assessed using the Japanese Orthopaedic Association shoulder score (JOA score) before the first injection, 1 week after the first injection, and 1 week after the final injection. Following each clinical evaluation, the patients underwent dynamic magnetic resonance imaging enhanced with Gd-DTPA, and the coefficient of enhancement (CE) in the glenohumeral synovium was calculated, with the examiners blinded to the clinical information. The JOA score tended to be greater and the CE smaller after injection than before injection. The changes in the CE following both single and repeated injections were negatively correlated with changes in the JOA score. Thus, clinical improvement in patients with FS was associated with a decrease in the CE. Because the CE depends on the degree of synovitis, the therapeutic effect of intraarticular HA injection for FS results, at least in part, from suppression of synovitis in the glenohumeral joint through an antiinflammatory effect. The Japanese Orthopaedic Association