The CURB-65 community-acquired pneumonia (CAP) severity score is a convenient 5-variable disease-specific tool validated to predict mortality and the decision to admit. This prospective study seeks to determine (a) the accuracy CAP diagnosis; (b) the degree to which the CURB-65 score was being documented; and (c) the degree to which CURB-65 was being utilised as an admission and early discharge tool in the A&E Department in the evaluation of patients presenting with CAP. Of 45 patients referred with an admitting diagnosis of CAP, 28 were subsequently found to satisfy the criteria of CAP, an over diagnosis of 38%. Documentation of CURB-65 was 7%; recording of core variables was 81% i.e. 4 out of 5 variables and confusion was the variable least likely to be documented. On re-scoring, 50% of patients had a score of 0 or 1 and a further 28% had a score of 2. We conclude that CAP is significantly overdiagnosed; that the CURB-65 severity score is not being utilised and that between 50 and 78% of patients with CAP may be admitted unnecessarily. This study raises important issues in the current management of CAP in the Irish Healthcare system.