There was a paucity of follow-up studies in the disease progression of early-onset PD patients with Parkin mutations (Parkin-EOPD). Here we conducted a longitudinal study to investigate the progression of motor and cognitive features of Parkin-EOPD patients. Genetic analysis was performed via target sequencing and multiplex ligation-dependent probe amplification. Thirty patients carrying homozygous or compound heterozygous Parkin mutations with at least 2 follow-up revisions were investigated as the Parkin-EOPD group. Fifty-two patients with at least 2 follow-up revisions, who did not have any known causative PD mutations, GBA or LRRK2 risk variants, a heterozygous Parkin mutation or 2 Parkin mutations without a segregation test, were defined as the genetically undefined EOPD (GU-EOPD) group. A linear mixed-effect model was implemented to evaluate longitudinal changes in motor symptoms and cognition. At baseline, the Parkin-EOPD group had a lower Unified Parkinson's Disease Rating Scale score (UPDRS-III) (off-medication) than the GU-EOPD group, without significant differences in cognition. A longitudinal study showed the estimated progression rate per year (standard error) of the UPDRS-III score (off-medication) was lower in the Parkin-EOPD group (0.203 [0.3162] points per year) than in the GU-EOPD group (1.056 [0.3001] points per year). The difference in the UPDRS-III score rate between the 2 groups was 0.853 (0.4183) (P = 0.042). The Parkin-EOPD group showed better maintenance of spatial processing ability compared with the GU-EOPD group (P = 0.027). Parkin-EOPD patients showed a slower deterioration of motor symptoms and a better spatial processing ability than GU-EOPD patients, which suggests that subtyping according to genetic features can help predict PD progression. © 2020 International Parkinson and Movement Disorder Society.