The consensus view that the amount of rock available for sampling does not significantly and systematically bias Phanerozoic marine diversity patterns has broken down. How changes in rock availability and sampling intensity affect our estimates of past biodiversity has been investigated with a variety of new approaches. A number of proxies for the amount of rock available for sampling have been used, but most of these proxies do not rely directly on evidence from large-scale geological maps (maps that cover small areas) and accompanying memoirs. Most previous map-based studies focused on single regions or relied on small-scale synoptic maps. We collected data from published geological maps and memoirs from western Europe, Australia, and Chile, which we combined with COSUNA data from the United States to generate the first multiregional data set for investigating whether the global Phanerozoic marine diversity record is a true global record, or is instead biased toward North America and Western Europe as has long been suspected. Both short and long-term trends in variation in the amount of outcrop display limited correlation among the regions studied. A series of diversification models obtained better matches to observed fossil diversity from the European and U.S. records than for the Chilean and Australian records, further supporting suspicions that the global Phanerozoic diversity curve is disproportionately influenced by European and U.S. fossil data. These results indicate that future research into Phanerozoic marine diversity patterns should not continue to apply global eustatic curves as a proxy for rock at outcrop, but should use regional data on rock occurrence.