A consecutive series of 205 myocardial scans, performed with 99mTc-labeled phosphates, in 185 patients with acute chest pain, were independently evaluated by comparing myocardial concentration intensity of the tracer to that in bone, and by rating this intensity by a six-category rating scale, which imply five criterion levels for calling an image "positive". The optimal criterion level was determined for each day of evolution of illness, as the one in which a shift to a more lax criterion level produces a bigger increment in false-positive results than in True-positive results, and in which a shift to a stricter criterion level causes a greater decrease in true positive results than in false-negative results. In all instances, the optimal criterion level was that with a moderate (2+) myocardial tracer concentration, lower than rib uptake, but with a focal pattern of myocardial distribution of the tracer (2F). By using this optimal criterion level, our results agree with the general consensus in that the procedure's usefullness is restricted to the first five days of evolution of illness. However, our daily statistical analysis showed that the best period to obtain the maximal diagnostic efficiency lies between the third and fourth days after the onset of acute chest pain. A negative result during this period, in patients with ECG and CPK serum concentration compatible with acute myocardial infarction, should be considered of prognostic significance, since it may translate a defficient colateral coronary circulation.