We prospectively examined 137 limbs in 112 consecutive patients with clinical evidence of severe varicosis by color coded duplex sonography and ascending venography (including varicography in 48 limbs) to evaluate the diagnostic capabilities of color coded duplex sonography in the assessment of venous anatomy, variant varicosis, postthrombotic changes, and incompetence of the superficial and perforating venous system. Additionally, descending venography was performed in the first 52 limbs and compared to color coded duplex sonography in the diagnosis of deep and superficial venous reflux. Variant venous anatomy (21 cases) was missed in two limbs and misinterpreted in one limb by ascending venography compared to surgery. Color coded duplex sonography was inconclusive in two cases. Variant varicosis (59 cases) was missed in seven surgically proved cases by venography and in one case by color coded duplex sonography. Color coded duplex sonography was inconclusive in five cases. Ascending venography was slightly superior to color coded duplex sonography in the detection of postphlebitic changes. Good agreement was found between color coded duplex sonography and descending venography in the grading of superficial (k = 0.75) and deep venous reflux (k = 0.79). Excellent agreement was found between ascending venography in the grading of long (k = 0.96) and short (k = 0.94) saphenous vein reflux. More incompetent perforating veins were detected by ascending venography, (and varicography) than by color coded duplex sonography, but the latter technique allows direct preoperative marking of the skin, which is beneficial for the surgeon. We conclude that color coded duplex sonography is a valuable imaging tool before venous stripping and is capable of replacing invasive ascending and descending venography. Only patients with inconclusive color coded duplex sonographic results (e.g., complex variant venous anatomy) should proceed to venography.