The evolutionary history of vascular plants is reviewed by extrapolation back through time from a wide range of data recently derived from the present flora, using as the central theme evolutionary inferences gained from phylogenies reconstructed as cladograms. Any region of the genome can be used to infer relationships, but only a combination of knowledge of morphology and the developmental genes that underpin morphology can allow evolutionary interpretation of macroevolutionary transitions; this in turn is necessary to identify bona fide evolutionary radiations and any putative causal key innovations. Such studies require clades to be delimited not by the inclusion of particular extant 'crown' species but rather by specific apomorphies, thereby giving important phylogenetic roles to extinct as well as extant species. Dating phylogenetic divergences via molecular clocks remains seriously inaccurate, and ultimately relies primarily on fossil benchmarks. First principles suggest that evolution of most regions of the genome is fundamentally gradual, whereas evolution of regions especially prone to strong selection pressure, and of the many facets of the phenotype, is punctuational, being characterized through time dominantly by stasis. Sequence data have proved valuable for inferring monophyletic groups, but within the now widely accepted context of monophyly the taxonomic hierarchy should primarily reflect degrees of morphological rather than molecular divergence. Incongruence among contrasting data sets is best explained by understanding the biological constraints operating on each type of phylogenetic information. The conventional 'uniformitarian' view of evolution has only limited applicability as one traces the history of land plants through time. Diversity increased in stepwise fashion, reflecting either attainments of complexity and/or fitness thresholds by the lineage (intrinsic) or the availability of unusually permissive environments, often following major perturbations (extrinsic). The Quaternary period demonstrates especially well the resilience, and ease of migration, of the Earth's vegetation. A higher frequency of generation of novel phenotypes in the deep past is possible, but a far higher frequency of their establishment is certain; together, these factors generate an evolutionary pattern of nested radiations that is fractal, as saturation of the resource space rendered the environment decreasingly permissive through time.In the immediate future, evolutionary-developmental genetics will have increasing value for testing homology, interpreting homoplasy and elucidating evolutionary constraints, and will become easierto pursue as whole-genome sequences of additional 'model' species further invigorate comparative genomics. Complexity of gene regulation, both by other genes and by the cellular and extra-cellular environment, appears a particularly fruitful area for further research. Nonetheless, environmental filtering of evolutionary novelties (whether instantaneously isolated mutant 'prospec...