From data on elemental concentrations in human tissues, Liebscher and Smith (1968) hypothesized that the frequency distributions of concentrations for essential elements were normal distributions, whereas the frequency distributions for nonessential elements were lognormal distributions. Although Liebscher and Smith's justifications for this hypothesis are flawed, other researchers have reported similar observations. We measured the concentrations of Cd, Ca, Cr, Co, Cu, Fe, Pb, Mn, Hg, K, Se, Na, and Zn in the muscle, liver, and egg tissues of female largemouth bass to determine (1) to what extent the frequency distributions of elemental concentrations are affected by errors in measuring concentrations and (2) whether the previously observed differences between essential and nonessential elements could be demonstrated if we restricted our comparisons to only those elements whose concentrations can be accurately measured. Variance component analyses of elemental concentrations in muscle tissue indicated that variations among replicate tissue samples due to measuring errors were large relative to the variations among individual fish for Cd, Ca, Mn, K, and Na. For elements where variation among individuals was not obscured by errors in measuring concentrations, there were no apparent differences between the frequency distributions for the essential elements, Cr, Cu,Fe, Se, and Zn, and the distributions for the nonessential elements, Pb and Hg. The hypothesis of Liebscher and Smith was not supported by our data.