Reliable identification of cyanobacterial isolates has significant socio-economic implications as many bloom-forming species affect the aesthetics and safety of drinking water, through the production of taste and odour compounds or toxic metabolites. The limitations of morphological identification have promoted the application of molecular tools, and encouraged the adoption of combined (polyphasic) approaches that include both microscopy- and DNA-based analyses. In this context, the rapid expansion of available sequence data is expected to allow increasingly reliable identification of cyanobacteria, and ultimately resolve current discrepancies between the two approaches.In the present study morphological and molecular characterisations of cyanobacterial isolates (n=39), collected from various freshwater sites in Australia, were compared. Sequences were obtained for the small ribosomal subunit RNA gene (16S rDNA) (n=36), the DNA-dependent RNA polymerase gene (rpoC1) (n=22), and the phycocyanin operon, with its intergenic spacer region (cpcBA-IGS) (n=19). Phylogenetic analyses identified three cyanobacterial orders: the Chroococcales (n=8), Oscillatoriales (n=6), and Nostocales (n=25). Interestingly, multiple novel genotypes were identified, with 22% of the strains (17/77) having <95% similarity to available sequences in GenBank.Morphological and molecular data were in agreement at the species level for only 26% of the isolates obtained (10/39), while agreement at the genus level was obtained for 31% (12/39). Confident identification of the remaining 44% of the strains (17/39) beyond the order level was not possible. The present study demonstrates that, despite the taxonomic revisions, and advances in molecular-, and bioinformatics-tools, the lack of reliable morphological features, culture-induced pleomorphism, and proportion of misidentified or poorly described sequences in GenBank, still represent significant factors, impeding the confident identification of cyanobacteria species.