TEX86L and TEX86H are organic palaeothermometers based on the lipids of Group 1 Crenarchaeota, recently proposed as a modified version of the original TEX86 index, but with significantly improved geographical coverage. Since few data from the global core top calibration are from the Pacific, this study was carried out to assess whether the global core top calibration is regionally biased or not. The result of principal components analysis of the fractional abundance of GDGTs, an analysis of variance (ANOVA) and the comparison of the residuals of TEX86H derived sea surface temperature (SST) estimates of the Pacific subset with that of the global data set suggest that the Pacific subset has a similar TEX86H-SST relationship with the global data set. However, the regression line through the Pacific data and an ANOVA on the residuals of TEX86L derived SST estimates suggest otherwise. The contradictory findings are likely to stem from the large scatter in the Pacific TEX86L values in the mid temperature range. While regionality does not seem to exert a strong bias on TEX86L and TEX86H calibration, it appears that there is a strong need to resolve the large scatter in the global data set, especially in the mid and high latitudes, in order to improve the calibration for a better SST estimation. © 2010 Elsevier Ltd.