Sebaceous wax ester secretion rates were measured in six subjects on six occasions each, using absorption into bentonite clay and analysis of the collected lipid by quantitative thin-layer chromatography. On each occasion eight samples were collected, four from the left and four from the right side of the forehead in four successive intervals. The first two intervals, which totaled 14 h, were intended to deplete the follicular reservoir of sebum so that a constant rate of absorption could be obtained during the third and fourth intervals, which were 3 h each. Thin-layer analysis of each sample was done in triplicate. The data were examined using analysis of variance techniques to determine the reproducibility of the measurement method and to identify possible sources of variability. The intraclass correlation coefficient (rI) for all 432 post-depletion determinations was 0.80. The reproducibility was considerably better for three of the subjects (rI = 0.93) than for the other three (rI = 0.75). Variability within the latter three subjects did not seem to be attributable to lack of reproducibility in the thin-layer analysis. Real biological variability also seems unlikely considering the holocrine mechanism of sebum secretion. Therefore, the variability probably arises from non-representative collection of sebum into the bentonite absorbent.