ObjectiveCompare efficacy of gene therapy alone (monotherapy) or in combination with an SMN2 augmentation agent (dual therapy) for treatment of children at risk for spinal muscular atrophy type 1.MethodsEighteen newborns with biallelic SMN1 deletions and two SMN2 copies were treated preemptively with monotherapy (n = 11) or dual therapy (n = 7) and followed for a median of 3 years. Primary outcomes were independent sitting and walking. Biomarkers were serial muscle ultrasonography (efficacy) and sensory action potentials (safety).ResultsGene therapy was administered by 7–43 postnatal days; dual therapy with risdiplam (n = 6) or nusinersen (n = 1) was started by 15–39 days. Among 18 children enrolled, 17 sat, 15 walked, and 44% had motor delay (i.e., delay or failure to achieve prespecified milestones). Those on dual therapy sat but did not walk at an earlier age. 91% of muscle ultrasounds conducted within 60 postnatal days were normal but by 3–61 months, 94% showed echogenicity and/or fasciculation of at least one muscle group; these changes were indistinguishable between monotherapy and dual therapy cohorts. Five children with three SMN2 copies were treated with monotherapy in parallel: all sat and walked on time and had normal muscle sonograms at all time points. No child on dual therapy experienced treatment‐associated adverse events. All 11 participants who completed sensory testing (including six on dual therapy) had intact sural sensory responses.InterpretationPreemptive dual therapy is well tolerated and may provide modest benefit for children at risk for severe spinal muscular atrophy but does not prevent widespread degenerative changes.