AbstractBackgroundThe current prevalence of X‐linked dystonia parkinsonism is not known. No screening tool for the condition exists. By developing a screening questionnaire, the present study is intended to be the first step in identifying the prevalence of the disease.AimTo develop and validate a simple, easy to use, community‐based, screening questionnaire for the diagnosis of X‐linked dystonia parkinsonism.MethodsCommunity health workers administered an 11‐item yes/no questionnaire in the native Panay island language to 54 genetically‐confirmed X‐linked dystonia parkinsonism patients and 54 healthy controls from the island of Panay. The questionnaire is made up of elements from existing questionnaires on Parkinson's disease and dystonia, and known clinical features of the disease. The participants were partitioned into training (n = 88) and test (n = 20) datasets. To select which items were predictive of X‐linked dystonia parkinsonism, the Clinical Utility Index of each item was determined. Afterwards, multivariable binary logistic regression was done to build a predictive model that was run on the test dataset.ResultsFour items on “sustained twisting,” “jaw opening and closing,” “slowness in movement” and “shuffling steps” were found to be the most predictive of X‐linked dystonia parkinsonism. All had at least a “good” Clinical Utility Index. The questions showed 100% sensitivity and 90% specificity (95% CI 65.6–100%) in identifying cases.ConclusionsThe resulting four‐item questionnaire was found to be predictive of X‐linked dystonia parkinsonism. The screening instrument can be used to screen for X‐linked dystonia parkinsonism in a prevalence study.