The diagnosis of pulmonary embolism (PE) and deep venous thrombosis (DVT) has long been based on purely clinical grounds. Pulmonary angiography and scintigraphy, as well as venography, reveal the poor diagnostic value of signs and symptoms of PE and DVT, which are now reduced to the role of triggers for these diagnostic tests. The development of simpler, but still imperfect diagnostic tools (mainly, D-climer plasma level, venous compression ultrasound, clinical probability of PE or DVT, spiral Cr pulmonary angiography) has led to the development of diagnostic strategies, which use combinations of tests, not to provide diagnostic certainty for every patient (presence or absence of clots), but rather to select populations of patients who need treatment and others who can safely remain untreated. Such pragmatic approaches must be validated in appropriate outcome studies. The choice of a diagnostic strategy should depend not only on the strategy's cost-effectiveness in the population under study but also on the local facilities and expertise required for its use.