Purpose: To examine the benefi ts of using near-isotropic single-slab three-dimensional (3D) magnetic resonance (MR) imaging for the creation of subtraction images and to evaluate their performance in the detection of active multiple sclerosis (MS) brain lesions in comparison with two-dimensional (2D) subtraction images. Materials and Methods: The study protocol was approved by the local ethics review board and all subjects gave written informed consent before investigation. Three-dimensional MR sequences, including double inversion-recovery, fl uid-attenuated inversion recovery, T2-weighted, and T1-weighted magnetization-prepared rapid acquisition gradient-echo (MP-RAGE), and corresponding 2D sequences were performed twice in 14 patients (eight women, six men; mean age, 37.6 years) with MS and nine age-matched healthy control subjects (three women, six men; mean age, 31.7 years). Active lesions were scored by two independent raters, followed by a consensus reading. Lesion counts were evaluated by using negative binomial regression; interrater agreement was evaluated by using intraclass correlation coefficient. Results: Three-dimensional subtraction images had less residual misregistration and fl ow artifacts and depicted higher numbers of active lesions with greater interobserver agreement compared with 2D subtraction images. Among the 3D sequences, MP-RAGE subtraction imaging enabled detection of a signifi cantly higher mean number of positive active MS lesions compared with 2D subtraction imaging (2.8 versus 1.7, P = .012), particularly infratentorial lesions (0.6 vs 0.1, P< .05), and a substantially higher (nonsignificant) mean number of small (< 3 mm) lesions (0.6 vs 0.1, P > .05). Conclusion: Three-dimensional subtraction imaging, after image registration, produced better image quality, leading to increased detection of active MS lesions with greater inter-observer agreement in comparison with 2D subtraction imaging; 3D MP-RAGE subtraction imaging represents a promising technique to increase sensitivity in ascertaining lesion dissemination in time and increase the power of MR imaging metrics for the evaluation of treatment effects in clinical trials. © RSNA, 2010.