OBJECTIVES: Walking impairment (WI) is a disabling hallmark of multiple sclerosis (MS). This patient-centric study aimed to assess the clinical, psychological, and economic burden of WI. METHODS: This cross-sectional study compared MS patients with (test group: TG, PDDS>0) and without (control group: CG,PDDS=0). WI as assessed by the Patient Determined Disease Steps (PDDS) questionnaire. PDDS=3 describes patients with mild WI while PDDS=4 reflects patients using a walking device. Using validated patient-reported outcomes instruments, falls, depression (Center for Epidemiologic Studies Depression Scale) resource utilization, and work productivity (Work Productivity and Activity Impairment Questionnaire) were assessed. Chi-square and t-test provided statistical significance to p<0.01. RESULTS: 346 adult MS patients around 11 years post-diagnosis participated. WI was significantly different across groups (TG:PDDS=3.2vs.CG:PDDS=0). In the TG, patients with WI using a device reported more healthcare visits for WI than patients with mild WI (PDDS=4:58%vs.PDDS=3:13%). TG patients reported “falling sometimes” significantly more than CG patients (PDDS=3:62%; PDDS=4:50% vs. PDDS=0:14%). Compared to CG patients, more PDDS=3 TG patients reported visiting the hospital/ER (11% vs. 5%) or doctor’s office (16%vs.5%) for falls. WI requiring the use of a device (PDDS=4) increased the reported healthcare visits (32%:hospital/ER;27%:doctor’s visits). TG patients reported depression significantly more than CG patients and 40% of these TG patients explicitly linked depression to WI. Absenteeism (10%vs.5%) and impairment while working (35%vs.18%) were reported more often in TG compared to CG patients. CONCLUSIONS: In this sample, WI was associated with increased falls, health care utilization, depression and lesser work productivity. WI requiring use of a device appears to increase these effects. WI effects could not fully be isolated from cognitive impairment and other effects, and confounding by indication could affect estimates. Multivariate model research is needed to validate these results in a broader population and quantify associated costs.