The diagnostic reliability and specificity of pulmonary angiography and perfusion photoscanning were examined by applying both technics in two series of dogs: one subjected to multiple small pulmonary emboli; the other, to massive embolization. Interpretations of the angiograms and photoscans by two observers were correlated with each other and autopsy data. Analysis disclosed a high degree of agreement between the two technics. Application of both technics, however, led to a greater incidence and accuracy of embolic detection than either method provided alone. Both technics were subject to interpretive errors leading to false positive and false negative diagnoses. The factors promoting such errors are discussed. It is concluded that the two technics are complementary, not competitive, and that each is least reliable in detecting emboli at the extremes of size.