To evaluate two published sets of prognostic classifications for patients with syncope, 176 consecutive patients who presented to an emergency room with syncope were studied. Although relatively few patients had cardiac syncope, these data confirmed their high one-year mortality. At the other extreme, it was also confirmed that patients who were 30 years of age or less or 70 years of age or less and had vasovagal/psychogenic syncope or syncope of unknown cause had a benign prognosis, with only two deaths in 225 patients in pooled data. However, these data did not confirm the previously reported prognoses for "medium-risk patients" or for patients with diagnosable noncardiovascular causes of syncope, largely because of differences in criteria for patient eligibility. It is concluded that available data allow over 70 percent of patients with syncope to be placed into either very-high or very-low-risk groups. However, further investigation, taking into account differences in patient selection criteria, will be required before accurate prognostic classifications can be derived for the nearly 30 percent of patients who do not fall into one of these extreme prognostic categories.