In this longitudinal study, the sensitivities of three magnetic resonance imaging techniques for detecting the appearance of new lesions in multiple sclerosis (MS) were evaluated and compared. Dual-echo conventional spin-echo (CSE), fast fluid-attenuated inversion recovery (fast-FLAIR) and post-contrast T1-weighted scans were obtained on four occasions, each separated by 28 days, from 18 patients with relapsing-remitting MS using a 1.5-T machine. New lesions seen using each sequence during the follow-up were counted by agreement by four observers in two stages (stage 1: random review of complete sets of scans from each technique; stage 2: side-by-side review with a 'retrospective' count of new lesions). At stage 1, 1.44 new lesions per patient per month were detected on CSE scans, 1.88 on fast-FLAIR (31% more than CSE) and 2.07 on post-contrast T1-weighted scans (44% more than CSE) (P = 0.03). Differences were, however, reduced after stage 2: fast-FLAIR detected 29% and post-contrast T1-weighted scans detected 31% more new lesions than CSE(P = 0.08). The combination of fast-FLAIR and post-contrast scans detected 144 new lesions, whilst the usual combination of CSE and post-contrast scans detected 133 new lesions. This study indicates that enhanced MRI remains the most sensitive method for detecting 'active' lesions in MS and that fast-FLAIR may be used when monitoring short-term disease activity in MS, either natural or modified by treatment.