Spinal muscular atrophy (SMA) is a neuromuscular disorder classified into four types based on the age of onset of the disease. Early onset is correlated with a higher mortality rate, mainly due to respiratory complications. Valproic acid (VPA) is a histone deacetylase (HDAC) inhibitor that has shown positive results on SMA both in experimental and cohort studies.This systematic review and meta-analysis aimed to investigate the efficacy and safety of VPA in patients with SMA.Eleven databases were systematically searched on 30 May 2017 for clinical trials that reported the efficacy and safety of VPA in SMA patients. The primary outcome was the efficacy of VPA in terms of gross motor function and expression of both full-length spinal motor neuron (SMN) gene (FL-SMN) and exon 7-lacking SMN. The secondary outcome was the safety of VPA in terms of reported adverse effects. The protocol was registered at PROSPERO (CRD42017067203).Five of the ten included studies were used in the meta-analysis (n = 126). The overall effect estimate, comparing pre- and post-VPA treatment, regardless of carnitine co-administration and design of the studies, showed significant improvement in gross motor function (standard mean difference [SMD] = 0.302, 95% confidence interval [CI] 0.048-0.556, P = 0.02) using the Hammersmith Functional Motor Scale (HFMS), Modified Hammersmith Functional Motor Scale (MHFMS), and MHFMS-Extend, with no significant heterogeneity. Similarly, in non-randomized controlled studies, the results indicated that there was a significant improvement detected (SMD = 0.335, 95% CI 0.041-0.628, P = 0.025), with no significant heterogeneity. Meanwhile, our results suggest that there was no significant improvement in treatment with co-administered carnitine (SMD = 0.28, 95% CI - 0.02 to 0.581, P = 0.067). No significant differences were found between pre- and post-VPA treatment co-administered with carnitine, in terms of the change in FL-SMN and exon 7-lacking SMN. Qualitative synthesis showed that other motor functions were not improved, while respiratory function test results were contradictory. Regarding the safety of the treatment, a double-blind, randomized, placebo-controlled trial reported no statistically significant differences for adverse events (AEs) between groups. Moreover, most of the included studies reported no serious AEs related to VPA use, although weight gain, gastrointestinal symptoms and respiratory symptoms were notable problems.Our study suggests that VPA treatment results in an improvement in gross motor functions for SMA patients, but not in other assessments of motor function or, possibly, in respiratory function. Furthermore, VPA appears to be a relatively safe drug, although treatment may be associated with a wide range of AEs (including body weight increase, fatigue, fever, flu-like symptoms, irritability, and pain). Double-blind, randomized, controlled trials are required to confirm these findings.