Three hundred and sixty samples of blood from 230 hospital patients were examined and compared with the results of simultaneous blood culture to determine the value of buffy-coat microscopy in detecting bacteraemia. One observer found 86 positive smears, 12 of which were from patients with positive blood cultures and 74 from patients with negative blood cultures. The buffy-coat smear was negative in 274 specimens, 8 of which yielded positive blood cultures. A second observer considered that only 34 of the buffy coats were positive. Only six of these were associated with positive blood cultures. Although there was a statistically significant association between positive smears and positive blood cultures, the procedure has little practical value because of the high incidence of false positives and negatives.