Parallel relative paleointensity records that are supposed to display the same geomagnetic variations often show discrepancies. We compared various techniques of normalization with different parameters involving alternating fields (a.f.) and thermal demagnetization of sediment samples spanning the same period from the Indian and the Pacific oceans. The results are basically unchanged using any technique. There is also no causal relationship between changes in rock magnetic properties and mismatch between the two records. Thus, discrepancies between parallel records do not depend on methods for relative paleointensity, and any technique can be used with confidence, provided that the characteristic component has been properly isolated and the sediments are magnetically homogeneous. Complications are more likely due to subtle variations in the sediment remanent magnetization acquisition. These results point to the importance of sedimentological (lithology and physical properties) factors in controlling the process of magnetization. ¿ 1998 American Geophysical Union