A statistical technique, components of variance analysis, was used to quantify three potential sources of variation in the determination of fluoride by ion chromatography (IC). The variance components due to sampling (2.09%) and oxygen bombings (1.47%) were found to be significant while the variance component due to IC repeats was only 0.272%. From these variance components the expected precision of various sampling and testing protocols was estimated. Graphical procedures were also used to visually analyze the experimental data.