The Bajocian to Oxfordian sequence of Gryphaea in Great Britain displays phyletic trends very similar to those exhibited by the famous and independent lineage of Liassic Gryphaea. In both lineages, shells become larger, relatively less coiled with reduced shell height, and more weakly sulcate. Larger, more saucer-shaped descendants are more stable than their ancestors; the sulcus, which enhanced stability in more strongly coiled forms, was no longer needed. Cretaceous Gryphaea may provide a third independent example of similar phyletic trends. We divide the Middle and Upper Jurassic sequence into four successive subspecies: G. bilobata bilobata (= G. sublobata) of the Lower Bajocian, G. bilobata calloviense subsp. nov. from the Lower Callovian, G. dilatata lituola from higher zones of the Callovian and G. dilatata dilatata from the Oxfordian. We took nine measures of basic size, coiling, attachment area and sulcus for these taxa and for four North American species living during the same time. In the bilobata-dilatata lineage, we separate evolutionary from ontogenetic trends on orthogonal axes of a varimax factor analysis. Phyletic trends towards increasing size and towards reduced height with weaker sulci can be distinguished. Centroids for the four subspecies project in temporal order upon the first axis of a discriminant analysis (76% of all information). When American species are added, all taxa separate reasonably well but we find no consistent contrast between the continents. Bivariate results distinguish the growth patterns leading to similar adult morphologies in the bilobata-dilatata as opposed to the Liassic lineage. Liassic Gryphaea exhibit strong ontogenetic allometry for increased coiling; evolution towards decreased coiling occurred by paedomorphosis affecting all stages of growth. The bilobata-dilatata lineage displays no ontogenetic allometry in post-juvenile growth; phyletic reduction in coiling can be traced to changes in early ontogeny alone. We summarize the stratigraphy and geographic distribution of Jurassic Gryphaea throughout the world. Gryphaea did not arise repeatedly by iterative evolution from a normal oyster stock. It first appears in the Upper Triassic, before Liostrea; Liassic Gryphaea can be derived from their Triassic congeners. We do not believe that Middle and Upper Jurassic Gryphaea arose from the ostreid Catinula. Moreover, the ecology of ostreids and gryphaeids seems to be consistently different throughout their history: ostreids tend to be opportunistic species inhabiting unpredictable environments within physically controlled communities of low diversity; gryphaeids are included in the ecosystems of predictable environments characterized by biologically accommodated communities of high diversity.