Traditional Giemsa-stained thick blood films were compared with 2 fluorescence microscopy techniques, acridine orange (AO) staining of thin blood films and the quantitative buffy coat (QBC) method, for the microscopical diagnosis of malaria. Of 200 samples examined, 141 were positive by Giemsa staining, 146 by AO and 137 by QBC. Overall sensitivities for the 2 fluorescence techniques compared to Giemsa staining were good: AO 97.9% and QBC 93.6%. However, with parasitaemias < 100/microL the QBC sensitivity fell to 41.7% whereas that of AO was 83.3%. Both AO and QBC were unable to differentiate accurately between individual malaria species. We conclude that the QBC technique alone cannot replace Giemsa-stained thick blood films for most purposes in an African setting. However, apart from species differentiation, the AO method is an appropriate technique for the laboratory diagnosis of malaria in developing countries.