The Silurian-Devonian transition is a critical moment in early land plant evolution. During this interval, plants invaded the continent, exploring different types of morphologies, and diversifying into multiple lineages, some of which have direct living representatives. One of these lineages, in particular, plays a central role in the understanding of these evolutionary events: the zosterophylls. This group reached highest diversity during the Early Devonian and is thought to include the ancestors of lycopsids. Found in floras all over the world, they are considered a dominant component of Early Devonian vegetation. Since the definition of the group in 1968 based on a small number of genera, numerous new fossils were discovered and today more than 40 genera of zosterophylls s.l. are known. They encompass a wide range of morphological diversity, a subset of which has been included in a formally defined Zosterophyllopsida. To date, two studies have probed with large datasets the relationships among zosterophylls and between the latter and other polysporangiophytes, obtaining conflicting results. Kenrick and Crane (1997) recovered the euphyllophytes sister to a clade within which Zosterophyllopsida and Lycopsida are sister to each other. Additionally, they recovered lycopsids in a polytomy with a Zosterophyllopsida clade and other zosterophylls s.l. In the other major analysis, Hao and Xue (2013) focusing on taxa of the Posongchong flora of China, recovered zosterophylls and lycopsids in separate clades that are part of a paraphyletic grade to the euphyllophytes,or a polytomy between these three main groups. Here, we present results based on the most comprehensive phylogenetic matrix of zosterophylls to date (56 taxa, 49 characters; parsimony-based). Results show a polytomy with rhyniophytes, protracheophytes, and a large clade including all other polysporangiophytes. Within this clade, zosterophylls are not monophyletic, because the lycopsids and euphyllophytes are nested among them. This is, thus, a third hypothesis of general phylogenetic relationships among early polysporangiophytes, based on a large dataset. Some of the conflict between the results of the three studies is due to differences in taxon sampling, especially with respect to the representation of the different main groups. However, conflict is certainly also resulting from homoplasy among the simple morphologies of different early polysporangiophytes, specifically since this increases with denser taxon sampling. This pattern is consistent with a rapid Early Devonian evolutionary radiation in several polysporangiophyte lineages. Our results question the monophyly of Zosterophyllopsida, suggesting that zosterophylls s.l. are stem group tracheophytes, and Lycopsida and Euphyllophytina the crown groups with extant representatives. Together, these results show that we have not reached consensus yet regarding the relationships among zosterophylls and between them and other groups of early tracheophytes. The current conflicting relationships may indicate that our exploration of the fossil record has yet to reach the minimum thresholds of taxon sampling density and character construction required for stable resolution of these major phylogenetic nodes.