The companion paper describes how implementation of professional networks (ParkinsonNet) may improve the quality and efficiency of allied health care in Parkinson's disease (PD). We designed a cluster-randomized controlled trial to evaluate this ParkinsonNet concept for one allied health discipline, namely physical therapy. Here we describe the study design and baseline characteristics. The design fully complies with the CONSORT criteria. Sixteen regions in the Netherlands were randomly divided into eight experimental regions where a ParkinsonNet was implemented, and eight control regions where the organization of care was left unchanged (usual care). Participating patients were followed for 6 months to evaluate the implementation process, health benefits and costs of the intervention. In the ParkinsonNet regions, 46 therapists were trained and 358 patients were included. In the usual care regions, 341 patients were included. Baseline characteristics of participants in the ParkinsonNet and control clusters were comparable. With 699 participating patients, this is the largest allied health study in PD to date.