Abstract
The measurement error for femoral oblique and total tibial length was examined based on data of the author and of Nishi (1928), using modern Japanese samples and Martin's method. Intra- and inter-observer measurement error was quite small for the femoral oblique length and the difference between the two variances was not significant in an F-test. Oblique length data in the literature may thus be considered reasonably accurate. For the total tibial length, inter-observer error was greater than intra-observer error and the difference between the two variances was significant in the F-test. Total tibial length may differ slightly between measurement values for the same sample obtained by different observers.
