Minor endothelial injury was produced in pial arterioles by a laser-Evans blue technique. This technique previously has been shown to prevent relaxation of the arterioles by topical acetylcholine (ACh) or bradykinin, agonists whose actions are known to be endothelium dependent in most vascular beds. L-Arginine is thought to be the precursor of the endothelium-derived relaxing factor (EDRF) that mediates the response to ACh. To further test this hypothesis and to begin exploration of the intracellular effects of the laser injury, we used the laser to eliminate the response to ACh and then attempted to restore the responsivity by suffusing the arterioles with 10(-3) L-arginine for 10 min before testing with ACh. This treatment restored the response. In contrast, over the same time period, responses failed to recover in arterioles that were never exposed to L-arginine or in arterioles that were only exposed to L-arginine at the time of retest with ACh. D-Arginine failed to restore the response to ACh. L-Arginine failed to restore the response to bradykinin. The data support the hypothesis that L-arginine acts as a substrate for synthesis of "classical" EDRF in brain microvessels in vivo. The data suggest that laser-Evans blue injury either depletes the endothelium of L-arginine or interferes with conversion of L-arginine to EDRF. Increasing the substrate by exposing the vessels to L-arginine for 10 min permitted ACh to release sufficient EDRF to elicit dilatation. The data also support conclusions from earlier studies that in this vascular bed the dilations produced by ACh and bradykinin are mediated by different EDRF.