In 2005, Hillson and colleagues developed a method for obtaining cervicometrics on a collection of loose teeth. However, I identified problems with the method when trying to apply it to material from archaeological collections with a mixture of in situ and loose teeth. The correct placement of the caliper tips at the suggested mesiodistal landmarks proved impossible for many in situ teeth. Their recommendation to rotate teeth to access the suggested landmarks resulted in large errors because not all teeth could be rotated. Other problems were identified with the suggested buccolingual dimensions of molars, which produced measurements that are not homologous across tooth class because of differential reduction in the distal cusps. Am J Phys Anthropol 154:159–164, 2014. © 2014 Wiley Periodicals, Inc.