Although previous work has demonstrated that biological phosphates (‘biophosphates’) record significant changes in δ18O associated with variations in local climate and seasonality, the repeatability of these analyses between laboratories has not previously been tested. We serially sampled enamel on four Cretaceous dinosaur teeth for phosphate δ18O analysis at up to three different facilities. With the exception of one set of unprocessed enamel samples, the material supplied to each laboratory was chemically processed to silver phosphate. Each laboratory analyzed sample sets by pyrolysis (thermochemical decomposition) in a ThermoFinnigan TC/EA attached to a ThermoFinnigan Delta Plus mass spectrometer. Significant interference between phosphate samples and the NIST reference material 8557 barium sulfate (NBS 127) distorts some of the results. Samples analyzed immediately following NBS 127 may be depleted by 6‰ isotopically and in instrument peak amplitude response by 80%. Substantial interference can persist over the subsequent 20 silver phosphate samples, and can influence the instrument peak amplitude response from some organic standards. Experiments using reagent-grade silver phosphate link these effects to divalent cations, particularly Ca2+ and Ba2+, which linger in the reactor and scavenge oxygen evolved from pyrolysis of subsequent samples. Unprocessed enamel includes 40 wt% calcium and self-scavenges oxygen, disrupting the isotopic measurements for the first half of a set and depleting subsequent organic standards by up to 9‰. In sets without NBS 127 or calcium, such interference did not occur and an interlaboratory comparison of results from enamel shows reproducible, significantly correlated peaked δ18O patterns with a 2–3‰ dynamic range, consistent with previous results from contemporaneous teeth. Whereas both unprocessed enamel and the NBS 127 barium sulfate should be applied to biological phosphate (‘biophosphate’) stable isotope research with caution, seasonal variations in enamel phosphate δ18O are a paleoecologically valuable, reproducible phenomenon in theropod dinosaur teeth. Copyright © 2004 John Wiley & Sons, Ltd.