This prospective, multicentre, international, observational, cohort study compared injection site pain (ISP) and injection site reactions (ISRS) between interferon beta-1b (IFNB-1b; Betaferon (R)) 250 mu g subcutaneously every other day and interferon beta-1a (IFNB-1a; Rebif (R)) 44 mu g subcutaneously three times weekly in patients with relapsing-remitting MS. Patients started treatment within 3 months before recruitment and were on full dose of therapy at inclusion. Patients self-injected IFNB and self-assessed ISP for 15 consecutive injections immediately, 30 and 60 min after injection, using a visual analogue scale diary. Study staff assessed ISRS. Of 445 patients (valid cases), similar to 90\% used autoinjectors. More patients were pain-free at all timepoints with IFNB-1b than with IFNB-1a (eg, 30 min: 42.6\% versus 19.7\%; P < 0.0001). The mean proportion of pain-free injections was greater for IFNB-1b (eg, 30 min: 79.0\%) than for IFNB-1a (53.3\%; P < 0.0001). The proportion of patients without ISRS was greater for IFNB-1b (second visit 51.8\% versus 33.8\%; P < 0.0001). Compared with IFNB-1a, more IFNB-1b patients either had no pain or their ISP had no influence on treatment satisfaction (76.9\% versus 64.1\%; P=0.006). The impact on tolerability and patient acceptability of any new IFNB product formulations would, however, have to be evaluated in comparative studies.