To appraise the evidence on compound muscle action potential (CMAP) as a biomarker in spinal muscular atrophy (SMA): responsiveness to therapy, correlation with motor function, diagnosis and prognostic value, and patterns across SMA types.

Conducted per PRISMA 2020 (PROSPERO CRD420261476823). PubMed and Embase were searched to August 15, 2026 for studies reporting CMAP in genetically confirmed SMA. Findings were synthesized narratively across four pre-specified domains using a Synthesis Without Meta-analysis approach. Two reviewers independently assessed risk of bias with Joanna Briggs Institute tools.

Sixty-two studies (2513 participants across reports) were included. Following disease-modifying therapy, CMAP increased in 22 of 33 contributing studies (66.7%), was stable in 9 (27.3%), and showed a mixed pattern in 2 (6.1%); none reported a net decrease. All three disease-modifying therapies were associated with CMAP increase. CMAP correlated moderately to strongly with motor function scales and discriminated diagnostic or severity categories in 6 of the 11 studies that formally tested this. Risk of bias was low or low-moderate in 27 studies; small sample size and absence of a concurrent control group were the most frequent limitations.

CMAP rises in roughly two-thirds of treated cohorts and reflects disease severity, but floor, ceiling, and site-dependent effects limit its use as a standalone surrogate endpoint. Selecting the recording site by severity, adding a second site chosen on the same basis, and agreeing standardized acquisition and normative values would strengthen its role as a prognostic stratifier and a component of multimodal models.