BackgroundPREDICT‐PD is a United Kingdom population‐based study aiming to stratify individuals for future Parkinson's disease (PD) using a risk algorithm.MethodsA randomly selected, representative sample of participants in PREDICT‐PD were examined using several motor assessments, including the motor section of the Movement Disorder Society‐Sponsored Revision of the Unified Parkinson's Disease Rating Scale (MDS‐UPDRS)‐III, at baseline (2012) and after an average of 6 years of follow‐up. We checked for new PD diagnoses in participants seen at baseline and examined the association between risk scores and incident sub‐threshold parkinsonism, motor decline (increasing ≥5 points in MDS‐UPDRS‐III) and single motor domains in the MDS‐UPDRS‐III. We replicated analyses in two independent datasets (Bruneck and Parkinson's Progression Markers Initiative [PPMI]).ResultsAfter 6 years of follow‐up, the PREDICT‐PD higher‐risk group (n = 33) had a greater motor decline compared with the lower‐risk group (n = 95) (30% vs. 12.5%, P = 0.031). Two participants (both considered higher risk at baseline) were given a diagnosis of PD during follow‐up, with motor signs emerging between 2 and 5 years before diagnosis. A meta‐analysis of data from PREDICT‐PD, Bruneck, and PPMI showed an association between PD risk estimates and incident sub‐threshold parkinsonism (odds ratio [OR], 2.01 [95% confidence interval (CI), 1.55–2.61]), as well as new onset bradykinesia (OR, 1.69 [95% CI, 1.33–2.16]) and action tremor (OR, 1.61 [95% CI, 1.30–1.98]).ConclusionsRisk estimates using the PREDICT‐PD algorithm were associated with the occurrence of sub‐threshold parkinsonism, including bradykinesia and action tremor. The algorithm could also identify individuals whose motor examination experience a decline over time. © 2023 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.