Purpose To explore the effects of customized 3D-printed assistive technology (AT) on functional performance and feasibility in patients with neurological impairment. Methods Patients with neurological impairment were recruited and randomized into customized 3D-printed assistive device group (group 1; n = 17) or standard device group (group 2; n = 14). The device was designed to assist their writing, spoon using, and typing. Each patient underwent 4-week intervention with the device (30 min per session, twice a week). Results We observed significant differences in shoulder abduction (p = .00), external rotation (p = .01), and internal rotation (p = .02) in group 1. And significant differences in abduction (p = .05) and external rotation (p = .05) between the 2 groups. Group 1 achieved significant improvements in writing without AT (p = .04) and with AT (p = .02), spoon use without AT (p = .02) and with AT (p = .03), and hemiplegia-side typing with AT (p = .00). Group 2 achieved significant improvements in writing without AT (p = .01), hemiplegia-side typing without AT (p = .01), and bil-side typing with AT (P = .05). Moreover, no significant differences were noted in other outcome measures. Conclusions This study demonstrated that customized 3D-printed AT can improve shoulder active motion for patients with neurological impairment. A positive effect in functional hand tasks after AT intervention. Offering customized AT with specific training could enhance the efficacy of interventions. The feasibility of using 3D printing technology to produce customized AT, which has the potential to be cost-effective and efficient.