The effect of different extraction techniques on the recovery of fatty acids from freeze-dried biomass of two lipid-producing microheterotrophs was examined. Two procedures were used: the extraction of lipids from biomass followed by transesterification of the fatty acids (extraction-transesterification); and the direct transesterification of biomass to produce fatty acid methyl esters (i.e. without the initial extraction step). Variable factors in the extraction-transesterification experiment were the sequence in which solvents were added to the samples, the relative amount of methanol in the solvent mix, and sonication of biomass while in the solvent mix. Variable factors in the direct transesterification experiment were sample size, and reaction duration. Statistical analysis of data (level of significance P<0.05) showed that: (1) extraction of total fatty acids prior to transesterification was significantly more efficient when solvents were added in the order of increasing polarity; (2) neither sonication nor increasing the proportion of methanol in the extraction solvent significantly affected extraction of fatty acids prior to transesterification; (3) efficiency of direct transesterification of fatty acids increased significantly with reaction time; (4) efficiency of direct transesterification of fatty acids was not significantly affected by sample size; (5) the most efficient method for extraction of fatty acids prior to transesterification yielded significantly less fatty acids than the most effective direct transesterification method. While the study examined only two strains, our results suggest that fatty acid analysis methodology for microheterotrophs under consideration for biotechnological exploitation requires optimisation and validation.