<p>Supplementary figures and tables.</p>

Two authors independently screened studies, with data extraction by one and verification by a second. We assessed risk of bias using a Joanna Briggs Institute critical appraisal tool. We conducted meta-analyses in RevMan for studies reporting age-sex-adjusted or multivariable-adjusted mortality ratios and R software to assess for publication bias.

We reported heterogeneity using the <i>I</i>^<i>2</i> and <i>τ</i>^<i>2</i> statistics. We identified 12,664 records of which 29 studies met the inclusion criteria, representing over 35 million children/adolescents. Sixteen studies were eligible to be included in the meta-analysis.

Overall, children with disabilities had a 5.31-fold (95% confidence interval (CI) [3.63,7.77]; <i>p</i> < 0.001; <i>τ</i>^<i>2</i> = 0.88; <i>I</i>² = 98%) higher all-cause mortality than their peers. The association between disability and mortality was higher in both younger children 0–5 years (Hazard ratio (HR)=7.51: 95%CI [4.70,12.00]; <i>p</i> < 0.001; <i>τ</i>^<i>2</i> = 0.44; <i>I</i>^<i>2</i> = 91%) and children 6–19 years (HR = 4.68: 95%CI [2.26,9.69]; <i>p</i> < 0.001; <i>τ</i>^<i>2</i> = 0.93; <i>I</i>^<i>2</i> = 96%).

The association varied by disability type and was most stark for children with neurological disabilities (HR = 11.71: 95%CI [7.90,17.35]; p < 0.001; <i>τ</i>^<i>2</i> = 0.19; <i>I</i>^<i>2</i> = 76%). Interpretation is limited by substantial between-study heterogeneity, differences in methods used to ascertain disability ascertainment and in the effect measures used, and the small number of studies from low-income and middle-income countries.</p> <p>Conclusions</p><p>Children with disabilities experience substantially higher mortality than their peers without disabilities.

There is an urgent need for evidence on what works to close the mortality gap.</p></div>