Both paleobiology and investigations of ‘major evolutionary transitions’ are intimately concerned with the macroevolutionary shape of life. It is surprising, then, how little paleontological perspectives and evidence inform studies of major transitions. I argue that this disconnect is partially justified because paleobiological investigation is typically ‘phenomena-led’, while investigations of major transitions (at least as commonly understood) are ‘theory-led’. The distinction turns on evidential relevance: in the former case, evidence is relevant in virtue of its relationship to some phenomena or hypotheses that concern those phenomena; in the latter, evidence is relevant in virtue of providing insights into, or tests of, an abstract body of theory. Because paleobiological data is by-and-large irrelevant to the theory which underwrites the traditional conception of major transitions, it is of limited use to that research program. I suggest that although the traditional conception of major transitions is neither ad-hoc or problematically incomplete, its promise of providing unificatory explanations of the transitions is unlikely to be kept. Further, examining paleobiological investigations of mass extinctions and organogenesis, I further argue that (1) whether or not transitions in paleobiology count as ‘major’ turns on how we conceive of major transitions (that is, the notion is sensitive to investigative context); (2) although major transitions potentially have a unified theoretical basis, recent developments suggest that investigations are becoming increasingly phenomena-led; (3) adopting phenomena-led investigations maximizes the evidence available to paleobiologists.