Background. Vagus nerve stimulation (VNS) paired with rehabilitation may improve upper-limb impairment and function after ischemic stroke. Objective. To report 1-year safety, feasibility, adherence, and outcome data from a home exercise program paired with VNS using long-term follow-up data from a randomized double-blind study of rehabilitation therapy paired with Active VNS (n = 8) or Control VNS (n = 9). Methods. All people were implanted with a VNS device and underwent 6 weeks in clinic therapy with Control or Active VNS followed by home exercises through day 90. Thereafter, participants and investigators were unblinded. The Control VNS group then received 6 weeks in-clinic Active VNS (Cross-VNS group). All participants then performed an individualized home exercise program with self-administered Active VNS. Data from this phase are reported here. Outcome measures were Fugl-Meyer Assessment—Upper Extremity (FMA-UE), Wolf Motor Function Test (Functional and Time), Box and Block Test, Nine-Hole Peg Test, Stroke Impact Scale, and Motor Activity Log. Results. There were no VNS treatment–related serious adverse events during the long-term therapy. Two participants discontinued prior to receiving the full crossover VNS. On average, participants performed 200 ± 63 home therapy sessions, representing device use on 57.4% of home exercise days available for each participant. Pooled analysis revealed that 1 year after randomization, the FMA-UE score increased by 9.2 points (95% CI = 4.7 to 13.7; P = .001; n = 15). Other functional measures were also improved at 1 year. Conclusions. VNS combined with rehabilitation is feasible, with good long-term adherence, and may improve arm function after ischemic stroke.