Waist-to-height ratio (WHtR) predicts abdominal fat and cardiometabolic risk. In children with obesity, the most adequate cut-off to predict cardiometabolic risk as well as its ability to predict risk changes over time has not been tested. Our aim was to define an appropriate WHtR cut-off to predict cardiometabolic risk in children with obesity, and to analyze its ability to predict changes in cardiometabolic risk over time. This is an observational prospective study secondary to the OBEMAT2.0 trial. We included data from 218 participants (8-15 years) who attended baseline and final visits (12 months later). The main outcome measure was a cardiometabolic risk score derived from blood pressure, lipoproteins, and HOMA index of insulin resistance. The optimal cut-off to predict the cardiometabolic risk score was WHtR ≥0.55 with an area under the curve of 0.675 (95% CI: 0.589-0.760) at baseline and 0.682 (95% CI: 0.585-0.779) at the final visit. Multivariate models for repeated measures showed that changes in cardiometabolic risk were significantly associated with changes in WHtR. This study confirms the clinical utility of WHtR to predict changes in cardiometabolic risk over time in children with obesity. The most accurate cut-off to predict cardiometabolic risk in children with obesity was WHtR ≥0.55. In children, there is no consensus on a unique WHtR cut-off to predict cardiometabolic risk. The present work provides sufficient evidence to support the use of the 0.55 boundary. We have a large sample of children with obesity, with whom we compared the previously proposed boundaries according to cardiometabolic risk, and we found the optimal WHtR cut-off to predict it. We also analyzed if a reduction in the WHtR was associated with an improvement in their cardiometabolic profile.