Objectives. To compare occurrence of injection site reactions (ISRs) in patients with relapsing remitting multiple sclerosis (RRMS) newly started on interferon beta 1b (Betaferon), using 3 delivery methods. Study design. A randomized, multicenter, phase IV, open label crossover study was performed in 82 sites in France on 294 patients with RRMS beginning a treatment with interferon beta 1b. For the first month all patients used a standard injection technique. They then used an autoinjector, Betaject. or Betaject. Light, for one month each, according to the cross-over design. Primary outcome was defined as the percentage of injections sites with ISR evaluated by the investigator. Secondary endpoints included graduation of ISR, using a five-point scale by both investigators and patients, injection related pain assessed by patients, percentage of patients without ISR and a global evaluation by patients of injection devices. Results. The percentage of ISRs were significantly reduced (p < 0.0001) when using either Betajector Betajectlight (24.1 percent and 24.1 percent respectively) compared with the standard technique (35.9 percent). No significant difference was seen between the 2 autoinjectors. The mean ISR intensity scores according to physician or patient were significantly reduced (p < 0.0001 for each) by the 2 autoinjectors compared to the standard injection technique. No significant difference on the pain scale comparing respectively the standard, Betajectand Betajectlight techniques but the mean level of pain was less than 1.2/10. In addition, the percentage of ISR free patients was significantly lower with the standard injection technique phase (52.4 percent) than with autoinjector use (respectively 68.1 and 66.7 percent). A non significant higher percentage of patients subjectively preferred Betaject(53.7 percent) than to Betajectlight (46.3 percent). The main other adverse events reported were flu-like symptoms (30.7 percent), transient and moderate increase of transaminases (4.8 percent)and headache (4.4 percent). Conclusion. We conclude that autoinjector use reduces the occurrence of ISR during IFNB-1b therapy in RRMS. Objectives. To compare occurrence of injection site reactions (ISRs) in patients with relapsing remitting multiple sclerosis (RRMS) newly started on interferon beta 1b (Betaferon), using 3 delivery methods. Study design. A randomized, multicenter, phase IV, open label crossover study was performed in 82 sites in France on 294 patients with RRMS beginning a treatment with interferon beta 1b. For the first month all patients used a standard injection technique. They then used an autoinjector, Betaject. or Betaject. Light, for one month each, according to the cross-over design. Primary outcome was defined as the percentage of injections sites with ISR evaluated by the investigator. Secondary endpoints included graduation of ISR, using a five-point scale by both investigators and patients, injection related pain by patients, percentage of patients without ISR and a global evaluation by patients of injection devices. Results. The percentage of ISRs were significantly reduced (p <0 .0001) when using either Betajector Betajectlight (24.1 percent and 24.1 percent respectively) compared with the standard technique (35.9 percent). No significant difference was seen between the 2 autoinjectors. The mean ISR intensity scores according to the physician or patient were significantly reduced (p <0.0001 for each) by the 2 autoinjectors compared to the standard injection technique. No significant difference on the pain scale Comparison of the standard, Betaject  and Betaject  light techniques but the mean level of pain was less than 1.2 / 10. In addition, the percentage of ISR free patients was significantly lower with the standard injection technique phase (52.4 percent) than with autoinjector use (respectively 68.1 and 66.7 percent). A non significant higher percentage of patients subjectively preferred Betaject (53.7 percent to than Betaject light (46.3 percent). The main other adverse events reported were flu-like symptoms (30.7 percent), transient and moderate incWe conclude that autoinjector use reduces the occurrence of ISR during IFNB-1b therapy in RRMS.