Abstract
Summary: A retrospective study of laboratory reports of sputums examined in the first 3 months of 1977 and of 1978 showed some interesting findings, which were similar in these 2 years. Analysis of the findings demonstrated that information from a proportion of sputum cultures were not helpful to the clinicians. Probable causes of such unhelpful results and some ways to overcome these problems were discussed, along with reviews of relevant literature. A conclusion drawn at the end of the discussion was that, under present circumstances, it was not expected that sputum cultures would produce totally reliable aetiological agents, unless attempts were made to obtain a better quality of specimen, e.g. those obtained by transtracheal or bronchoscopic aspirations.