For patients suspected of having arterial injuries form penetrating trauma to the extremities, the decision to perform arteriography should be based on clinical criteria rather than "proximity" alone. In order to demonstrate the increased yield of arteriography resulting from appropriate patient selection, we reviewed the results of 500 arteriograms performed at the West Side Veterans Affairs Medical Center to exclude arterial injuries in symptomatic patients following penetrating trauma to their extremities. According to their clinical presentation, the patients were grouped into three different screening categories A, B, and C, which correspond respectively to "mild," "soft," and "hard" clinical signs and symptoms suggestive of arterial injury. All symptomatic patients injured by shotgun were included in category C due to the weapon's destructive power. The majority (318/500, 63.6%) of the 500 arteriograms were positive, and 102 of the positive studies identified major injuries that required surgical intervention. The remaining 216 identified injuries were minor or self-limiting. Increasing incidence of major injuries was found when comparing patients with mild (1.22%), soft (27.8%), and hard (60.0%) clinical signs and symptoms of arterial injury. These result suggest that arteriography, when combined with appropriate patient selection based on clinical examination, would be most effective in finding significant arterial injuries.