Spinal muscular atrophy (SMA), one of the most frequent and devastating genetic disorders causing neuromuscular degeneration, has reached the forefront of clinical translation. The quite unique genetic situation of SMA patients, who lack functional SMN1 but carry the misspliced SMN2 copy gene, creates the possibility of correcting SMN2 splicing by antisense oligonucleotides or drugs. Both strategies showed impressive results in pre-clinical trials and are now in Phase II-III clinical trials. SMN gene therapy approaches using AAV9-SMN vectors are also highly promising and have entered a Phase I clinical trial. However, careful analysis of SMA animal models and patients has revealed some limitations that need to be taken very seriously, including: i) a limited time-window for successful therapy delivery, making neonatal screening of SMA mandatory; ii) multi-organ impairment, requiring systemic delivery of therapies; and iii) a potential need for combined therapies that both increase SMN levels and target pathways that preserve/rescue motor neuron function over the lifespan. Meeting these challenges will likely be crucial to cure SMA, instead of only ameliorating symptoms, particularly in its most severe form. This review discusses therapies currently in clinical trials, the hopes for SMA therapy, and the potential limitations of these new approaches.