Objectives The aim of this study was to improve medication safety for children; it is important to quantify the occurrence of preventable medication errors (MEs). A trigger tool may be an effective and time-saving strategy, but its measurement performance is unclear. Therefore, we aimed to estimate the performance of a pediatric medication-focused trigger tool in detecting harmful MEs. Methods First, we established a multifaceted method as a reference comparison. Second, we compared the pediatric medication-focused trigger tool with the multifaceted method in a new cohort of patients. All patients admitted in February and March 2013 were screened using the trigger tool and the multifaceted method to obtain full verification. Data collection was performed in separate teams to guarantee blinding of the test results. Results Review of the clinical records and the voluntary incident reports was most effective in detecting harmful MEs, so this approach was chosen as a reference comparison. In the second part of the study, 369 patients were included. The multifaceted method identified 33 harmful MEs. In contrast, the trigger tool did not identify any harm. When the 2 symptoms pain and nausea/vomiting were added to the trigger tool, 19 harmful MEs were identified. This extended trigger tool resulted in a sensitivity of 21.2 and a positive predictive value of 36.8. Conclusions The original pediatric medication-focused trigger tool yielded only false-positive scores and left unsafe situations undiscovered. We conclude that a multifaceted method remains the preferred method to detect harmful MEs. The additional value of the trigger tool stays unclear.