Oral and pharyngeal swallowing dysfunction are common complications in acute stroke patients. This primary aim of this study was to determine whether oral neuromuscular training improves swallowing function in participants with swallowing dysfunction after stroke. A secondary aim was to assess how well results of the timed water-swallow test (TWST) correspond with swallowing dysfunction diagnosed by videofluoroscopy (VFS).This was an intention-to-treat two-centre prospective randomized open-label study with blinded-evaluators (PROBE) design. At 4 weeks after stroke onset, participants with swallowing dysfunction were randomized to 5 weeks of continued orofacial sensory-vibration stimulation with an electric toothbrush or additional oral neuromuscular training with an oral device (Muppy®). Participants were examined with TWST, a lip-force test, and VFS before (baseline), after 5 weeks' treatment (the end-of-treatment), and 12 months after treatment (follow-up). The baseline VFS results were compared with the TWST results. The primary endpoint was changes in swallowing rate assessed using TWST, from baseline to the end of training and from baseline to follow-up based on intention-to-treat analyses. The secondary endpoint was the corresponding changes in lip-force between baseline, the end of treatment, and follow-up.The participants were randomly assigned as controls (n = 20) or for intervention with oral neuromuscular training (n = 20). After treatment, both groups had improved significantly (intervention, P < 0.001; controls, P = 0.001) in TWST but there was no significant between-group difference in swallowing rate. At the 12-month follow-up, the intervention group had improved further whereas the controls had deteriorated, and there were significant between-group differences in swallowing rate (P = 0.032) and lip force (P = 0.001). A TWST < 10 mL/sec at baseline corresponded to VFS-verified swallowing dysfunction in all assessed participants.The 5-week oral neuromuscular training improved swallowing function in participants with post-stroke dysphagia compared with the controls 12 months after intervention, but there was no between-group difference in improvement immediately after treatment. TWST results corresponded with VFS results, making TWST a feasible method for identifying persons with swallowing dysfunction after stroke. Larger randomized controlled trials are required to confirm our preliminary positive long-term results.Retrospectively registered at ClinicalTrials.gov : NCT04164420. Registered on 15 November 2019.