Pulmonary embolism is a common and potentially lethal disease. Given the variable presentation and associated morbidity of this condition, an accurate and efficient diagnostic algorithm is required. Clinical pretest probability serves as the root of any diagnostic approach. We, thus, review several clinical decision rules that may help standardize this determination. Using a review of the literature, the accuracy, predictive values, and likelihood ratios for several diagnostic tests are described. The combination of these tests, based on the pretest probability of disease, can be used in a Bayesian fashion to make accurate treatment decisions. A completely noninvasive diagnostic algorithm for patients presenting with suspected acute pulmonary embolism is proposed.