MicroRNAs (miRNAs) play an important role in cellular differentiation and cancer pathogenesis. This study analysed the expression of 154 human miRNAs in acute myeloid leukaemia (AML) and control samples using a stem-loop real-time reverse transcription polymerase chain reaction approach. Global patterns of miRNA expression in AML, normal bone marrow (NBM) and CD34(+) progenitor cells allowed correct class predictions similar to whole genome microarray expression analyses that were performed at the same time. At single miRNA species level, MIRN23B was repressed in AML specimens compared to NBM and purified CD34(+) haematopoietic progenitor cells. In contrast, the MIRN221/MIRN222 cluster and MIRN34A were expressed at significantly higher levels in AML blasts. Patients with high MIRN221/MIRN222 expression showed low levels of KIT RNA and protein expression but the correlation between kit protein and KIT mRNA was significantly stronger than the correlation of either one with MIRN221/MIRN222. A global analysis between miRNA expression levels and mRNA expression of predicted target genes revealed only weak associations in the majority of miRNA species. Nonetheless, the presence of two or more miRNA binding sites within the mRNA was usually associated with a decrease in mRNA levels. Taken together, these findings provide evidence that specific miRNA expression patterns exist in AML.