Recently, a new anthropometric method of accurate lower leg length measurement (knemometry) was introduced. The present study was performed to investigate rigorously its strengths and weaknesses for auxological purposes, and to promote a standardized method of use. The data are based on 6 x 2200 single estimations of the lower leg length in 90 children of both sexes with normal, tall or short stature aged between 2.4 and 17.1 years. The overall technical error of this measurement was found to be 0.16 mm, which is reduced to 0.13 mm if the first estimation within each series of six is ignored. The following sources of systematic error could be detected: intra-daily variation, dependency on physical pressure put on or removed from the leg, vigorous physical activity prior to the measuring procedure, and a slow but significant correlation between day-to-day variation of lower leg length and body weight (r = 0.299, b = 0.372 mm/kg). There was no covariation with atmospheric parameters such as outside temperature, barometric pressure and others. After elimination of a linear trend from the individual growth curves, there was a remaining day-to-day variation of SD = 0.697 mm. This represents real changes of lower leg length and cannot be explained by the technical error of measurement which contributes to less than 10% of the total variation. Some of this day-to-day variation may be explained by non-linearity of growth. Knemometry appears to be a robust and precise new tool for the investigation of short term kinetics of longitudinal growth. Proposals for a standardized use of this technique are given.