Phylogenetic relationships of Equisetum are firmly established within the euphyllophytes, but more detailed resolution of sphenopsid phylogeny has not been achieved. Traditional paleobotanical studies recognize five groups of sphenopsids including Hyeniales, Pseudoborniales, Calamitales, Sphenophyllales and Equisetales, but these hypothesized relationships have not been fully tested using cladistic methodology. Sphenopsids plus cladoxylalean/zygopterid "ferns" resolved as the sister group to lignophytes in an analysis of euphyllophytes by Rothwell (1999), but Equisetum nested among the ferns as the sister to marattialeans in the recent analysis of living plants by Pryer et al. (2001). Relationships among sphenopsid groups are equally uncertain. Whereas Equisetales and Calamitales are widely regarded as closely related, their relationships to Pseudobornia, Hyenia and Sphenophyllum are far less certain. Indeed, Stewart (1980) has hypothesized that Sphenophyllum may be more closely related to lycophytes than to euphyllophytes. To test these competing hypotheses, we have undertaken a numerical cladistic analysis of representative rhyniophytes, zosterophylls, lycophytes, and euphyllophytes, including all of the hypothesized groups of sphenopsids. Relationships of these taxa are resolved using a matrix of more than 100 morphological characters. Principal questions addressed by this analysis include the following. Do the sphenopsids (as traditionally circumscribed) form a clade? What are the sister group relationships of the sphenopsid clade that includes Equisetum? And, are Sphenophyllales euphyllophytes, or do they nest within the zosterophyll/lycophyte clade?

Key words: cladistics, Equisetum, phylogeny, Sphenophyllales, sphenopsids