The diagnostic approach to deep vein thrombosis (DVT) has evolved during the last 3 decades. Contrast venography has been replaced by noninvasive tests. Compression ultrasonography (CUS) is currently the most widely used diagnostic test. Whereas CUS has a high accuracy for proximal DVT (thrombosis of the popliteal and more proximal veins), it has been shown to lack sensitivity and specificity for distal DVT. Ultrasonography can either be limited to the proximal veins and repeated within 1 week (serial limited CUS) or extended to both proximal and distal veins and performed on one occasion (single complete CUS). Both strategies are reliable diagnostic options for the management of patients with suspected DVT. The main limitation of proximal CUS is the need to repeat the test once in patients with initial negative findings. Conversely, complete CUS detects many distal DVTs for which systematic anticoagulation therapy is debatable and exposes patients to potentially unnecessary anticoagulation. Incorporation of D-dimer testing and clinical pretest probability assessment in the diagnostic algorithm is beneficial because it allows excluding DVT without the need for diagnostic imaging in about a third of patients.