Parkinson’s disease Multimodal Complex Treatment (PD-MCT) is a multidisciplinary inpatient treatment approach that has been demonstrated to improve motor function and quality of life in patients with Parkinson’s disease (PD). In this study, we assessed the efficacy of PD-MCT and calculated predictors for improvement. We performed a prospective analysis in a non-randomized, open-label observational patient cohort. Study examinations were done at baseline (BL), at discharge after two-weeks of inpatient treatment (DC) and at a six-week follow-up examination (FU). Besides Movement Disorders Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) III as a primary outcome, motor performance was measured by the Timed Up-and-Go (TUG), the Berg Balance Scale (BBS) and the Perdue Pegboard Test (PPT). Until DC, motor performance improved significantly in several parameters and was largely maintained until FU (MDS-UPDRS III BL-to-DC: −4.7 ± 1.2 (SE) p = 0.0012, BL-to-FU: −6.1 ± 1.3 p = 0.0001; TUG BL-to-DC: −2.5 ± 0.9 p = 0.015, BL-to-FU: 2.4 ± 0.9 p = 0.027; BBS BL-to-DC: 2.4 ± 0.7 p = 0.003, BL-to-FU: 1.3 ± 0.7 p = 0.176, PPT BL-to-DC: 3.0 ± 0.5 p = 0.000004, BL-to-FU: 1.7 ± 0.7 p = 0.059). Overall, nontremor items were more therapy responsive than tremor items. Motor complications evaluated with MDS-UPDRS IV occurred significantly less frequent at DC (−1.8 ± 0.5 p = 0.002). Predictor analyses revealed an influence of initial motor impairment and disease severity on the treatment response in different motor aspects. In summary, we demonstrate a significant positive treatment effect of PD-MCT on motor function of PD patients which can be maintained in several parameters for an extended time period of six weeks and identify predictors for an improvement of motor function.