Background: Valbenazine (VBZ, NBI-98854) is a novel vesicular monoamine transporter 2 (VMAT2) inhibitor approved for the treatment of tardive dyskinesia (TD). The KINECT 3 study (NCT02274558) evaluated the effects of VBZ on TD in subjects with schizophrenia/schizoaffective disorder (SCHZ) or mood disorder (mood disorder presented separately) who received up to 48 weeks of treatment., Methods: KINECT 3 included: 6-week, double-blind, placebo (PBO)-controlled (DBPC) period (205 completers); 42-week VBZ extension (VE) period (124 completers): 4-week washout period (121 completers). Subjects entering the DBPC were randomized 1:1:1 to once-daily VBZ 80 mg, VBZ 40 mg, or PBO; stable concomitant antipsychotic medication regimens were allowed. Subjects completing the DBPC and entering the VE period were re-randomized (blinded) 1:1 from PBO to VBZ (80 or 40 mg) or continued VBZ treatment at the same dose. Efficacy assessments included: mean changes from baseline in Abnormal Involuntary Movement Scale (AIMS) total score (items 1-7); mean Clinical Global Impression of Change (CGI-TD) scores; AIMS responders (subjects with ≥50% score reduction from baseline): and CGI-TD responders (subjects with score ≤2 ["much improved" or "very much improved"]). Treatment effect sizes (Cohen's d) and numbers needed to treat (NNTs) were analyzed for DBPC outcomes., Results: Efficacy analyses were conducted in 148 subjects (DBPC) and 125 subjects (VE) with SCHZ. At Week 6 (end of DBPC), AIMS mean score improvements were greater in the VBZ groups (in a dose-related pattern) than in the PBO group (80 mg, -2.9, d = 0.88; 40 mg, -1.6, d = 0.52; PBO, +0.3). AIMS score changes at Week 48 (end of VE) showed continued TD improvement during long-term VBZ treatment (80 mg, -4.2; 40 mg, -2.5). By Week 52 (end of washout), AIMS scores were returning toward baseline levels, indicating re-emergence of TD. CGI-TD mean scores were as follows: Week 6 (80 mg, 3.0, d = 0.11; 40 mg, 2.9, d = 0.23; PBO, 3.2), Week 48 (80 mg, 2.2; 40 mg, 2.4), Week 52 (80 mg, 3.4; 40 mg, 3.3). AIMS responder rates (≥50% score reduction) were greater with VBZ than with PBO at Week 6 (80 mg, 40.9%, NNT = 4; 40 mg, 26.2%, NNT = 6; PBO, 9.3%), were increased at Week 48 (80 mg, 50.0%; 40 mg, 26.2%), and decreased after VBZ washout (80 mg, 21.6%; 40 mg, 9.5%). CGI-TD responder rates followed a similar pattern: Week 6 (80 mg, 29.5%, NNT = 17; 40 mg, 33.3%, NNT = 10; PBO, 23.3%), Week 48 (80 mg, 73.7%; 40 mg, 58.1%), Week 52 (80 mg, 29.7%; 40 mg, 33.3%)., Conclusion: Sustained TD improvements were found in subjects with SCHZ who received up to 48 weeks of VBZ, with TD reverting toward baseline when assessed 4 weeks after treatment withdrawal. Together with results from mood disorder subjects and the long-term safety profile (presented separately), these results indicate that long-term VBZ can be beneficial for managing TD regardless of psychiatric diagnosis category.