Objective Our objective in this study is to evaluate whether brain magnetic resonance imaging (MRI) performed at interferon-beta (IFN-β) onset and after 12 months allow us to identify relapsing–remitting multiple sclerosis (RRMS) patients with a disability increase in the first 2 years of therapy. Methods This is a prospective and longitudinal study of patients with RRMS treated with IFN-β. All patients included underwent brain MRI before the onset of therapy with IFN-β and 12 months after. MRI measures (T2, unenhanced T1-weighted and gadolinium-enhancing T1-weighted brain lesion load, brain parenchymal fraction) were undertaken at baseline and after 12 months. The number of active lesions (new or enlarging T2 plus gadolinium-enhancing brain lesions) was also assessed on the 12 months MRI scan. Expanded Disability Status Scale (EDSS) was scored every 3 months. We defined an increase in disability as an increase of at least 1 EDSS point confirmed and sustained during the first 2 years of therapy with IFN-β. Regression analysis was performed in order to identify MRI variables of response. Results We included 152 patients who were followed-up for at least 2 years. After 2 years of therapy, 24 patients (16%) had an increase in disability. The logistic regression model showed that active lesions in the scan performed at 12 months were the most important factor related with the increase of disability after 2 years of therapy (odds ratio 8.3, 95% confidence interval 3.1–21.9; p < 0.0001). Conclusions In RRMS patients treated with IFN-β the MRI changes occurring during the first year may have a prognostic value for identifying patients with a confirmed increase of disability after 2 years of therapy.