Background: Acute limb ischemia represents an emergency in which restoration of perfusion through early intervention can lead to limb salvage, whereas delayed intervention may result in significant morbidity, including limb loss and potentially, death.Purpose: To prove the accuracy and reliability of duplex scanning in differentiating embolic from thrombotic acute arterial occlusion, and to test the role of post-operative duplex in confirming the pre-operative findings.Patients and methods: We prospectively recruited 58 patients; with 63 non-traumatic acute limb ischemia in native arteries. Contrast angiography or surgery served as the reference standard. We measured the arterial diameter at the site of occlusion (dO) and at the corresponding contralateral healthy side (dC). The change (∆) & fractional change (∆%) between the two diameters were calculated as: (dO - dC) and (∆ / dC) x 100 respectively. According to the reference standard, limbs were classified into embolic (E-group = 38 limbs) & thrombotic (T-group = 25 limbs) groups. Postoperative duplex study was done in 34 patients after embolectomy and the change (∆P) & fractional change (∆P%) between the postoperative (dP) & preoperative (dO) diameters at the site of occlusion were calculated as: (dP - dO) & (∆P / dO) x 100 respectively. We recruited 30 healthy volunteers and we measured the diameter of the superficial femoral artery just after its origin and the popliteal artery at the popliteal crease in both limbs by a single operator and then comparison was done between both sides. Also we recruited 20 healthy volunteers and we measured the diameter of the superficial femoral artery after its origin by two different operators to assess the interobserver reliability. Results: The mean age of the patients was 52.9 ± 16.5 years, 52.4% of them were females. We calculated (∆%) as 26.8 ± 39.3 % in the E-group vs. -11 ± 16.2 % in the T-group, (p<0.001). A cutoff value of ≥ 6.78% for (∆%) had 92% sensitivity and 84% specificity for the diagnosis of embolic vs. thrombotic occlusion (CI 0.86 to 0.99, p<0.001). It is more precise to use (∆%) rather than (∆) as it allows for a more unified measurement for comparison between arterial segments of different diameters. The (∆%) was the most important predictor for embolic and thrombotic occlusion. Postoperatively (∆P%) was -11.8% ± 8.2% with a statistically significant negative correlation was found between (∆) & (∆P), Spearman′s coefficient (rho) = -0.912, P<0.001. There was no statistically significant difference between the diameters measured at the same level bilaterally which validates using the contralateral segment as the reference segment. Interobserver agreement yielded an ICC value of 0.99 (95% [CI]: 0.977 - 0.996). Conclusions: In acute peripheral arterial occlusion, a cut off value of 6.78% as a fractional dilatation or diminution in the diameter of occluded artery is the most important duplex sign for predicting embolic or thrombotic occlusion respectively using the contralateral segment as the reference segment. Postoperative reduction in the diameter of occluded artery after embolectomy confirms this sign. Diameter measurements on B-mode images are largely observer independent with a very good interobserver agreement.