Definitive diagnosis of Creutzfeldt Jakob disease (CJD) remains tissue-based. Possible and probable CJD are useful clinical terms but may be used indiscriminately. The aim of this study was to assess the effectiveness of the Irish surveillance system and to ascertain how diagnostic accuracy in identifying clinically "definite" cases might be improved. We reviewed the clinical information, relevant investigations, and samples n = 100; (autopsy n = 87; biopsy n = 13) in 96 patients between January 1, 2005 and December 31, 2015. In 4 cases both a biopsy and autopsy were performed. CJD was confirmed in 50 patients (45 at autopsy and 5 at biopsy). Sporadic CJD (sCJD) accounted for 90% of cases (n = 45); variant CJD (vCJD) for 6% (n = 3) with 1 case each of familial CJD and iatrogenic CJD. CSF 14-3-3, EEG, and MRI investigations were helpful but not available on all patients. CJD mimics (n = 46) fell into the following categories: neurodegenerative (n = 22), immune mediated (n = 3), cerebrovascular disease (n = 5), tumor (n = 5), dual pathology (n = 3), and miscellaneous (n = 8). The Irish surveillance system fulfils its main objective as all clinically suspicious cases are being referred. CJD was confirmed in 52% (n = 50/96) of referrals. Based on this, we propose an algorithm for CJD referrals to reduce both infection control and diagnostic difficulties encountered in CJD surveillance. .