Although archaeomagnetic dating seems straightforward in principle, there are practical limitations which are not generally understood. Unlike rate-dependent processes such as isotopic dating, archaeomagnetic dating requires the construction of a master record of geomagnetic secular variation. Error is inherent in such a master curve due to statistical uncertainties regarding both the magnetic directions and ages of the samples used to create the curve. The master curve itself is thus best represented as a ribbon rather than a line. Features being dated have their own error of measurement of magnetic direction, and deriving a date involves an interpretation based on the relation between the oval of confidence for the unknown and the ribbon representing the master curve. Thus a practical precision limit for archaeomagnetic dating is about ± 20 years under optimal circumstances, but the limit will generally be higher. Our pilot study revealed no major discrepancies between our work and the curves of DuBois (1975).