We conducted a meta-analytic review of neuropsychological studies of mild head trauma (MHT). Studies were included if they met these criteria: patients studied at least 3 months after MHT; patients selected because of a history of MHT rather than because they were symptomatic; and attrition rate of less than 50% for longitudinal studies. Studies of children were not considered. We found a total of 8 published papers with 11 samples that met these criteria. Using the g statistics, the overall effect size of 0.07 was nonsignificant, but the d statistic yielded an effect size of 0.12, p < .03. Measurers of attention had the largest effect, g = 0.17. p < .02 and d = 0.20, p < .006. Severity of injury accounted for far more variance than did specific neuropsychological domain, however. The small effect size suggests that the maximum prevalence of persistent neuropsychological deficit is likely to be small and neuropsychological assessment is likely to have positive predictive value of less than 50%. Consequently, clinicians will more likely be correct when not diagnosing brain injury than when diagnosing a brain injury in cases with chronic disability after MHT.