Difficulties with mobility and cognition are common in multiple sclerosis (MS). This study used a dual-task (DT) paradigm to examine long-term DT learning and neuropsychological function in MS. Sixty participants (30 adults with MS and 30 control participants) attended three study sessions conducted 1 month apart at Holy Name Medical Center in New Jersey. Measures included single-task gait speed, DT gait speed (walking while counting backwards), and the Minimal Assessment of Cognitive Function in MS. Results indicated significant main effects of learning over the first 30 and total 60 days, task type, and disease status, which were modified by an interaction between task type and disease status, F(1, 202.86) = 11.63, p = .001. Regarding neuropsychological functioning, significant interactions included: visit and BVMT-R TR, F(2, 221.28) = 3.74, p = .025; visit and BVMT-R DR, F(2, 219.33) = 4.14,p = .017; task, disease status, and COWAT FAS, F(1, 195.19) = 4.03, p = .046; disease status and CVLT-II LDFR, F(l, 50.85) = 4.45,p = .040 (adjusted model); visit and DST FS, F(2, 220.74) = 3.21, p = .042; visit, task, disease status, and JLO, F(2, 230.94) = 3.96, p = .020; and visit, task, disease status, and PASAT, F(2, 233.66) = 3.07, p = .048. Significant main effects of learning over the first 30 and total 60 days, task type, and SDMT were modified by an interaction between task and disease status, F(1, 202.68) = 11.65,p = .001. Overall, both groups of participants learned over the first 30 days and total 60 days, without significant differences between groups. Visual learning/memory, working memory/processing speed, executive functioning, and visuospatial judgment played roles in long-term learning of DT gait speed. This is likely the first non-interventional study to compare long-term learning of DT gait speed in MS and control groups, and to evaluate whether neuropsychological function moderates the effect of DT learning in MS. As a sensitive and ecologically valid measure that correlates with outcomes such as fall risk and pharmacological efficacy, long-term DT learning can add to existing assessments of cognitive-motor performance and functional ability, and rehabilitative tools.