There are a large number of microRNA (miRNA) prediction programs available using a range of strategies to predict and discover miRNA. In this study I examine a range of different bioinformatics tools to determine the efficiency and quality of miRNA predictions. This would enable researchers to more easily determine the best programs to select for the analysis of sRNA data identifying likely miRNA. For my comparison miRDeep-P, miRDEEP2, miRCat, miREAP, HHMMIR, miRFinder, ShortStack and PIPmiR were selected. These programs were compared on their ability to predict known and novel miRNA from an Arabidopsis thaliana smallRNA (sRNA) dataset. To determine the number of false positive sequences sRNA data from Dicer mutants were used. Through this analysis it is possible to determine the best approach for the determination of either conserved or novel miRNA. In addition, it can be determined if a wide range of approaches will increase the range and diversity of miRNA predicted or if combining strategies increases confidence in miRNA predictions. This information enables more directed approaches to determine miRNA within complex plant genomes.