Traumatic brain injury (TBI) is the leading cause of acquired disability in children, and attention-deficit/hyperactivity disorder (ADHD) is one of the most common neurodevelopmental sources of disability. 1,2 The association between TBI and ADHD has been a topic of controversy, which Asarnow et al 3 address in their systematic review and metaanalysis in this issue of JAMA Pediatrics. They focus on whether the risk of ADHD increases after TBI, but also present data bearing on the question of whether ADHD is itself a risk factor for TBI. Importantly, their analysis shows a "dose-related" association between TBI severity and postinjury ADHD, with a significantly elevated risk of ADHD among children with severe TBI compared with an other injured control (OIC) group of children and children with mild TBI within 1 year postinjury, and with a noninjured control (NIC) group, OIC, and children with mild TBI more than 1 year postinjury. In addition, their analyses show no increase in postinjury ADHD in children with concussion, mild TBI, or moderate TBI compared with the rate in the NICs or OICs. They also note that the rate of preinjury ADHD was higher in children with TBI than in the general population, although not higher than among samples of the NIC or OIC groups. These findings have important clinical implications, highlighting the need to assess for ADHD in children with TBI, especially those with severe injuries, but also to take a careful history to determine whether symptoms of ADHD predate the injury.The Asarnow et al study 3 has several strengths. These include the largest number of studies included in a systematic review on this topic to date (n = 24) and a substantial number of participants, including 12 374 unique patients with TBI and 43 491 unique controls. Another strength is the modern bayesian meta-regression analysis, which considers prior probabilities in an explicit fashion, as well as the sensitivity analyses, which addressed potential concerns about pooled severity groups for TBI samples, the association with less informative prior distributions, and whether the results were related more to specific studies. One minor shortcoming is that the estimates in the meta-analysis do not take diagnostic method for ADHD into account. The sensitivity analyses suggest that diagnostic method can have a substantial influence on estimates of ADHD prevalence. A specific comparison of estimates using standard structured psychiatric interviews with those using symptom rating scales or unstructured clinician diagnoses would have been informative. Another minor shortcoming was the absence of a specific test of the doseresponse relationship between severity of TBI and ADHD. The inclusion of a specific linear contrast for this purpose would have been instructive.