Abstract Background Drug screening by immunoassay is common in pediatric populations. However, false-positive and -negative results due to antibody cross-reactivity and dilute urine are frequent and underappreciated. Accurate ascertainment of drug exposure in children has significant clinical and medico-legal consequences. Design and Methods We developed and characterized an LC–MS/MS drug screening assay to supplant immunoassay and detect 38 compounds at the lowest concentrations distinguishable from analytic noise. Once implemented, we conducted a retrospective analysis of 3985 pediatric urine drug screens performed a year before (n = 1663) and after (n = 2322) implementation to examine the frequency and breadth of drug detection in our pediatric population. Results Using immunoassay, 23% (293/1269) of samples from the general pediatric and 37% (147/394) of nursery populations had presumptively positive results. Of the presumptive positive compounds, 85% (288/338) from the general pediatric population and 40% (65/162) from the nursery cohort were confirmed by mass spectrometry. After LC–MS/MS implementation, 31% (628/2052) of general pediatric, and 18% (48/270) of the nursery samples were positive for 1 or more compounds. In the nursery population, immunoassays over-detected the presence of THC but under-detected exposure to cocaine. Conclusion A broadly targeted, analytically sensitive LC–MS/MS drug screening assay detects a larger number and variety of compounds in a single step compared to a screen-then-confirm approach initiated by immunoassay in our pediatric population. Rapid delivery of accurate results enables timely, appropriate disposition of patients in a variety of settings including the emergency department and labor/delivery.