Upper extremity (UE) impairments in infants with cerebral palsy (CP) result from reduced quality of motor experiences and "noisy" sensory inputs. We hypothesized that a neuroscience-based multi-component intervention would improve somatosensory processing and motor measures of more-affected (UEs) in infants with CP and asymmetric UE neurologic impairments, while remaining safe for less-affected UEs. Our randomized controlled trial compared infants (6-24 months) with CP receiving intervention (N = 37) versus a waitlisted group (N = 36). Treatment effects tested a direct measurement of reach smoothness (3D-kinematics), a measure of unimanual fine motor function (Bayley unimanual fine motor raw scores), and EEG measures of cortical somatosensory processing. The four-week therapist-directed, parent-administered intervention included daily (1) bimanual play; (2) less-affected UE wearing soft-constraint (6 h/day, electronically-monitored); (3) reach training on more-affected UE; (4) graduated motor-sensory training; and (5) parent education. Waitlist infants received only bimanual play. Effectiveness and safety were tested; z-scores from 54 posttest-matched typically-developing infants provided benchmarks for treatment effects. Intervention and waitlist infants had no pretest differences. Median weekly constraint wear was 38 h; parent-treatment fidelity averaged > 92%. On the more affected side, the intervention significantly increased smoothness of reach (Cohen's d = - 0.90; p < .001) and unimanual fine motor skill (d = 0.35; p = .004). Using unadjusted p values, intervention improved somatosensory processing (d = 0.53; p = .04). All intervention effects referenced well to typically developing children. Safety of the intervention was demonstrated through positive- or non-effects on measurements involving the constrained, less-affected UE and gross motor function; unexpected treatment effects on reach smoothness occurred in less-affected UEs (d = - 0.85; p = .01). This large clinical trial demonstrated intervention effectiveness and safety for developing sensory and motor systems with improvements in reach smoothness, and developmental abilities.Clinical Trail Registration: ClinicalTrials.gov NCT02567630, registered October 5, 2015.