A new generation ⁸⁷Sr/⁸⁶Sr isotope curve for the Phanerozoic seawater, based on a compilation of 3635 stratigraphically well defined and well preserved skeletal components, results in a substantially narrower band, with some discrepancies, if compared to the 'curve' of Burke et al. (1982), which is based mostly on whole-rock samples. The results obtained from low-Mg calcite shells (foraminifers, belemnites, brachiopods and oysters) are mostly superior to those from phosphatic skeletal components, such as conodonts, even if the latter are exceptionally well preserved (CAI ≤1.5). At this stage, the Phanerozoic data set based on fossils is patchy for the Cambrian, Upper Ordovician-Lower Devonian and Upper Carboniferous-Triassic intervals. The major practical constraint that precludes delineation of an unequivocal Sr isotope curve for the Phanerozoic seawater is the uncertainty in the assignment of absolute ages to the samples. As a result the curve is in reality a band. This is a reflection of the limitations of biostratigraphy and geochronology that are inherited by any derivative correlation and dating technique, including isotope stratigraphy. It is therefore unlikely that, in the near future, the Sr isotope stratigraphy will surpass the resolution capability of biostratigraphy as a global correlation tool. On the other hand, if complemented by lithostratigraphy, it is potentially a valuable tool for correlation of sequences on intrabasinal and regional scales. Taking into account the above limitations, careful consideration must be given to assignment of higher order isotope peaks to specific 'events' within a biozone and to the use of such events as a global correlation tool.