BackgroundParkinson’s disease impacts health-related quality of life (HRQoL), however no studies inquired on predictors of HRQoL changes after rehabilitation. This study assessed the relationship between mobility domain of HRQoL measured by Parkinson’s Disease Questionnaires-39 (PDQ-39) and clinical-demographic characteristics and developed a model predicting changes after rehabilitation.MethodsSubjects with Parkinson’s disease underwent rehabilitation treatment and completed the following predictors: 10-m walking test (10MWT), Timed Up and Go (TUG), Berg Balance scale (BBS), Activities-specific Balance Confidence scales (ABC), Freezing of Gait (FOGQ) and PDQ-39. Two general linear models were calculated to predict the relationship between HRQoL at baseline and to predict HRQoL changes after rehabilitation.ResultsForty-two subjects (age 74.9 ± 7.3 years, Hoehn&Yahr 2.8 ± 0.6) completed the baseline evaluation. The first model (multiple R2 = 0.59, F = 5.86, P < 0.001) showed that ABC (B = − 0.51, CI = − 0.86 to 0.15, R2 = 0.41, P = 0.005) and FOGQ (B = 2.38, CI = 1.03 to 3.73, R2 = 0.07, P = 0.001) were statistically significant predictors of mobility aspect of HRQoL at baseline. Thirty seven subjects completed the rehabilitation sessions, data were entered in the second model (multiple R2 = 0.40, F = 4.24, P < 0.004) showing that gender (B = − 5.12, CI = − 9.86 to − 0.39, R2 = 0.23, P = 0.034), Hoehn&Yahr (B = 10.93, CI = + 3.27 to + 18.61, R2 = 0.22, P = 0.006) and PDQ-39 mobility at baseline (B = − 0.38, CI = − 0.63 to − 0.14, R2 = 0.55, P = 0.002) were statistically significant predictors of changes of the mobility aspect of HRQoL.ConclusionsBalance confidence and Freezing of Gait are associated with the mobility aspect of HRQoL. Changes in mobility domain of HRQoL (as assessed by PDQ-39) are likely to be greater in males, in people at higher stages of the disease and in people with more severe limitation in mobility domain of HRQoL (as assessed by PDQ-39) before rehabilitation. Results might be different when considering different outcomes or different measures for the same outcome (performance mobility test instead of self-report questionnaires). Further investigations are needed to better understand other components of HRQoL in addition to mobility.Trial registrationNCT02713971 registered March 8, 2016.