Spinal muscular atrophy (SMA) is a motor neuron disease and the leading genetic cause of infant mortality. SMA results from insufficient survival motor neuron (SMN) protein due to alternative splicing. Antisense oligonucleotides, gene therapy and splicing modifiers recently received FDA approval. Although severe SMA transgenic mouse models have been beneficial for testing therapeutic efficacy, models mimicking milder cases that manifest post-infancy have proven challenging to develop. We established a titratable model of mild and moderate SMA using the splicing compound NVS-SM2. Administration for 30 d prevented development of the SMA phenotype in severe SMA mice, which typically show rapid weakness and succumb by postnatal day 11. Furthermore, administration at day eight resulted in phenotypic recovery. Remarkably, acute dosing limited to the first 3 d of life significantly enhanced survival in two severe SMA mice models, easing the burden on neonates and demonstrating the compound as suitable for evaluation of follow-on therapies without potential drug–drug interactions. This pharmacologically tunable SMA model represents a useful tool to investigate cellular and molecular pathogenesis at different stages of disease.