Aims The most common reason of childhood obtructive sleep apnea (OSA) is adenotonsillar hypertrophy (ATH). In adults OSA may lead to alterations in P-wave duration. This study evaluated P-wave dispersion (PWD) in children undergoing adenotonsillar surgery as compared with a control group.Methods We retrospectively analysed data of 51 children who underwent adenotonsillar surgery (group AT) and compared with matched the control group (group C). We reviewed 12-lead ECGs of the children. Duration of maximum (Pmax) and minimum P-wave (Pmin) was recorded, and PWD is calculated. Lateral radiographs of the nasopharynx were evaluated and the measurement of adenoid to nasopharynx ratio (ANR) performed.Results PWD was significantly higher in group AT (41 ± 15 ms) than the control group (27 ± 8 ms) (p < 0.001). There was no correlation between PWD and the severity of upper airway obstruction assessed by adenoid to nasopharynx ratio. The cut-off value of PWD for symptomatic ATH children was 32.5 ms with a sensitivity of 69%, a specificity of 77%, a positive predictive value of 69% and an area under the curve of 0.789 (0.703–0.874, 95%CI).Conclusions PWD, a simple ECG marker, is significantly increased in children with symptomatic ATH. We suggest that a PWD longer than 32.5 ms in children scheduled for ATH surgery should prompt echocardiographic and polysomnographic investigations to evaluate the severity of OSA.Clinical trial number: not applicable.