ABSTRACTBackgroundIn patients with Parkinson's disease (PD), sleep, mood, cognitive, autonomic, and other non‐motor symptoms may fluctuate in a manner similar to motor symptoms.ObjectivesTo validate a final version of a patient‐rated questionnaire that captures the presence and severity of non‐motor fluctuations in levodopa‐treated PD patients (NoMoFA).MethodsWe recruited PD subjects from five movement disorders centers across the US and Canada. We assessed the internal consistency, floor and ceiling effects, test–retest reliability, and concurrent validity of NoMoFA. Classical test theory and item response theory methods informed item reduction and Delphi process yielded a final questionnaire.ResultsTwo hundred subjects and their care‐partners participated in the study (age: 66.4 ± 9.6 years; disease duration: 9 ± 5.5 years; median Hoehn and Yahr [H&Y] OFF: 3 [range 1−5]; mean Unified Parkinson's Disease Rating Scale (UPDRS) III ON score: 27.4 ± 14.9). Acceptability of the scale was adequate. There were floor effects in 8/28 items. Cronbach's alpha was 0.894. While eight items had “item‐to‐total” correlations below the cutoff of 0.4, removing these items did not improve Cronbach's alpha. Test–retest reliability was acceptable (intraclass correlation coefficient [ICC] 0.73; 95% confidence interval, 0.64−0.80). Concurrent validity was adequate with all Spearman's rho values comparing NoMoFA score to other measures of parkinsonian severity showing significance and in the expected direction. A final Delphi panel eliminated one item to avoid redundancy.ConclusionsThe final 27‐item self‐administered NoMoFA is a valid and reliable questionnaire, capturing both static and fluctuating non‐motor symptoms in PD. © 2021 International Parkinson and Movement Disorder Society