The heterocytous cyanobacteria form a monophyletic group according to 16S rRNA gene sequence data. Within this group, phylogenetic and morphological studies have shown that genera such as Anabaena and Aphanizomenon are intermixed. Moreover, the phylogeny of the genus Trichormus, which was recently separated from Anabaena, has not been investigated. The aim was to study the taxonomy of the genera Anabaena, Aphanizomenon, Nostoc and Trichormus belonging to the family Nostocaceae (subsection IV.I) by morphological and phylogenetic analyses of 16S rRNA gene, rpoB and rbcLX sequences. New strains were isolated to avoid identification problems caused by morphological changes of strains during cultivation. Morphological and phylogenetic data showed that benthic and planktic Anabaena strains were intermixed. In addition, the present study confirmed that Anabaena and Aphanizomenon strains were not monophyletic, as previously demonstrated. The evolutionary distances between the strains indicated that the planktic Anabaena and Aphanizomenon strains as well as five benthic Anabaena strains in cluster 1 could be assigned to a single genus. On the basis of the 16S rRNA, rpoB and rbcLX gene sequences, the Anabaena/Aphanizomenon strains (cluster 1) were divided into nine supported subclusters which could also be separated morphologically, and which therefore might represent different species. Trichormus strains were morphologically and phylogenetically heterogeneous and did not form a monophyletic cluster. These Trichormus strains, which were representatives of three distinct species, might actually belong to three genera according to the evolutionary distances. Nostoc strains were also heterogeneous and seemed to form a monophyletic cluster, which may contain more than one genus. It was found that certain morphological features were stable and could be used to separate different phylogenetic clusters. For example, the width and the length of akinetes were useful features for classification of the Anabaena/Aphanizomenon strains in cluster 1. This morphological and phylogenetic study with fresh isolates showed that the current classification of these anabaenoid genera needs to be revised.