We evaluated the ability of noninvasive vascular tests to exclude clinically significant occult arterial damage in injured extremities. In a preliminary study, a Doppler arterial pressure index (API) (the systolic AP in the injured extremity divided by the AP in an uninvolved arm) of less than 0.90 was found to have sensitivity and specificity of 95% and 97%, respectively, for major arterial injury. The negative predictive value for an API greater than 0.90 was 99%. Because these values suggested that noninvasive vascular tests might effectively be substituted for "exclusion" arteriography in patients at risk for silent extremity arterial injuries, we then conducted a trail in which arteriography was performed in extremity trauma victims only when the API was less than 0.90. Among 100 traumatized limbs (84 penetrating, 16 blunt) in 96 consecutive patients, 16 of 17 limbs (94%) with an API less than 0.90 had positive arteriographic findings, and seven underwent arterial reconstruction. Among 83 limbs with an API greater than 0.90, followup (including duplex scanning in 64 limbs) revealed five minor arterial lesions (four pseudoaneurysms, one arteriovenous fistula) but no major injuries. Arteriograms for extremity trauma fell from 14% to 5.2% of all angiographic studies performed (p less than 0.001, Chi-square). These studies suggest that noninvasive vascular tests can reliably exclude major occult arterial damage in injured extremities. Screening for such injuries with Doppler arterial pressure measurements, reserving arteriography for limbs in which the API is less than 0.90, is safe, accurate, and cost effective.