The benefit of selective dorsal rhizotomies (SDR) on motor function relative to the cerebral palsy (CP) natural history remains unknown. To determine the functional benefit of SDR over the longitudinal CP natural history. Retrospective, single-center, case-control study of patients post-SDR after 1990. Inclusion criteria were the following: diagnosis of spastic CP, at least 1 preoperative and 1 postoperative Gross Motor Function Measure (GMFM-88), at least 1 yr of postoperative follow-up. GMFM-88 assessments were performed at 1, 2, 3, 5, 10, and 15 yr postoperatively and converted to GMFM-66. Cases were stratified by preoperative Gross Motor Function Classification System (GMFCS) and matched against their expected natural history using published reference centiles. After age 12, our cohort and matched controls were also fitted to published nonlinear mixed models of GMFM-66 evolution over time. Analysis included 190 patients. Median follow-up, 5.3 yr (range: 1-16.9), median age at surgery, 4.6 yr, and 81.6% of patients grouped as GMFCS II or III pre-op. SDR patients performed statistically significantly better than their expected natural history (P < .0005). At 21 yr old, a modeled benefit of 8.435 was observed for GMFCS I (P = .0051), 0.05 for GMFCS II (P = .9647), 6.31 for GMFCS III (P < .0001), and 1.191 for GMFCS IV patients (P = .0207). These results support the use of SDR in carefully selected spastic CP children. Copyright © 2019 by the Congress of Neurological Surgeons.