Case ascertainment for an epidemiologic study of Creutzfeldt-Jakob disease (CJD) can be difficult. This report investigates the efficiency of various sources of case ascertainment for CJD. Cases were identified utilizing neuropathologists, hospitals and death certificates from 11 targeted states. For the period of 1986–1988, 247 death certificates indicating a diagnosis of CJD were obtained. Only 26 potential cases were identified without death certificates. The proportion of neuropathologically confirmed cases identified by death certificates only, i.e., which were not identified through any other source, was 42%. Furthermore, 80% of all the neuropathologically confirmed cases were ascertained utilizing death certificates as a source. Of the remaining 20%, 7% were ascertained through neuropathologists only, 10% through hospitals only, 1.5% through a combination of hospitals and neuropathologists, and 1.5% through another source. The false-positive rate for death certificates with neuropathology (which may have been performed after the death certificate was filled out) was estimated to be 8.3%. The results indicate that death certificates were by far the most efficient source for initial ascertainment of potential CJD cases to be followed by verification of diagnosis.