Abstract
In screening for anonymous mycobacteria the use of thiosemicarbazone (lO[mu]g/ml.) may give rise to false positive results. Although these can usually be overcome by further tests, additional time and effort is required, which seriously limits the value of thiosemicarbazone as a true screening agent. Para-nitrobenzoic acid (500 [mu]g./ml.) has proved more satisfactory in our hands; no false positive results have been obtained and it has the additional advantage that strains of M. kansasii (usually thiosemicarbazone sensitive) are resistant to para-nitrobenzoic acid.