Free radicals have been implicated in the etiology of many neurodegenerative conditions. Yet, because these species are highly reactive and thus short-lived it has been difficult to test these hypotheses. We adapted a method in which hydroxyl radicals are trapped by salicylate in vivo, resulting in the stable and quantifiable products, 2,3-dihydroxybenzoic acid (DHBA) and 2,5-DHBA. After systemic (100 mg/kg i.p.) or intraventricular (4 mumol) administration of salicylate, the amount of DHBA in striatal tissue correlated with tissue levels of salicylate. After systemic salicylate, the ratio of total DHBA to salicylate in neostriatum was at least 10-fold higher than that observed after central salicylate. In addition, systemic salicylate resulted in considerably higher concentrations of 2,3- and 2,5-DHBA in plasma than in brain. Therefore, a large portion of the DHBA present in brain after systemic salicylate may have been formed in the periphery. A neurotoxic regimen of methamphetamine increased the concentration of DHBA in neostriatum after either central or systemic administration of salicylate. The increase in 2,3-DHBA after the central administration of salicylate was significant at 2 h, but not at 4 h, after the last dose of methamphetamine. These results suggest that (1) when assessing specific events in brain, it is preferable to administer salicylate centrally, and (2) neurotoxic doses of methamphetamine increase the hydroxyl radical content in brain in a time-dependent manner.