Abstract The south Mediterranean region, including western Sicily, Crete and mainland Greece (southern Peloponnese and Evia), is critical to an interpretation of the Late Palaeozoic-Early Mesozoic tectonic evolution of Tethys. Several contrasting tectonic models compete to explain the regional evolution. In a divergence-related hypothesis (Model 1) the south Aegean region experienced pulsed rifting along the northern margin of Gondwana that culminated in break-up to form the Pindos ocean in the region of Greece. In an alternative convergence-related hypothesis (Model 2) the south Aegean experienced Late Palaeozoic Early Mesozoic northward subduction, accretion and arc magmatism, culminating in ‘Cimmerian’ suturing of a Palaeotethyan ocean in latest Triassic time. In a third model, southward subduction of a Palaeotethyan ocean took place beneath the North Gondwana margin during Late Palaeozoic-Triassic time, giving rise to back-arc magmatism in an extensional setting. In addition, a more complex setting involving two opposing subduction zones (Andean-type and intra-oceanic) has also been suggested (Model 4), mainly based on lava geochemistry. To test these tectonic alternatives, mainly sedimentary studies were carried out in western Sicily, western and eastern Crete, the Peloponnese and Evia (eastern central Greece). Western Sicily was studied as a proxy for the unexposed deep Mediterranean south of Crete. Most of the available evidence supports the divergence-related (pulsed rift) hypothesis (Model 1). There is no clear evidence of sea-floor spreading (e.g. ophiolites) to the south of what became the Pindos ocean, or of plate convergence (e.g. magmatic arcs, subduction complexes), or collisional deformation in the south Aegean region that could be related to subduction or collision during the Mid-Carboniferous to Triassic, as in Model 2. Model 3 is not supported by evidence from the wider region (northern Greece, Turkey). Model 4 is not supported by evidence independent of igneous geochemistry. In the proposed interpretation, the northern margin of Gondwana initially rifted during Mid-Carboniferous to Early Permian time to form a wide deep-water basin. This was followed by further rifting, associated with volcanism during the Early Triassic; final continental break-up and spreading to form the Pindos ocean to the north during Late Triassic to Early Jurassic time then followed. Mid-Triassic uplift of part of the rift basin is explained as a flexural response to rifting as a precursor to opening of the Pindos ocean. Passive margin subsidence during the Early Mesozoic relates to opening of the Pindos ocean to the north. A subduction geochemical signature within some Triassic volcanic rocks, in this interpretation, is explained by melting of heterogeneous sub-crustal mantle, following an earlier, possibly Hercynian, subduction event.