Two 140 year-old and two 75 year-old bigleaf maple (Acer macrophyllum Pursh)   and western hemlock (Tsuga heterophylla (Raf.) 20 x 20 m plots, centered on individual dominant stems, were sampled at 100 systematic locations and tested for forest floor pH, nitrate, and ammonium. The kriging approach was used to explore overall spatial patterns, spatial autocorrelation analysis (SSA) was used to detect global spatial patterns, LISA statistics located points of statistical significance, and variance partitioning was used to reveal the spatial component contribution to over all forest floor variability. The spatial structure of forest floor pH was detected by all of the analytical methods and was more pronounced on the 140 year-old plots. Krig maps showed that forest floor pH had higher and lower values in locations adjacent to the bigleaf maple and western hemlock stem respectively. SSA detected an overall positive spatial autocorrelation up to average distance of about 10 m and 7 m on bigleaf maple and western hemlock plots respectively. LISA showed the presence of positive cluster and negative cluster associated with bigleaf maple and western hemlock stems respectively, up to a distance of 3 m. Variance partitioning revealed that the spatial component has a higher contribution to overall variability on bigleaf maple plots. The spatial signal of nitrate and ammonium was relatively weak. SSA detected weak global spatial autocorrelation on bigleaf maple plots and random patterns on western hemlock plots. LISA showed very few clusters on most of the plots and to a distance not exceeding 2 m from the stem. Variance partitioning showed that environmental and spatial components contributed almost equally to the explained variation.  This study demonstrates that the spatial influence is specific to the examined soil property and that it becomes more structured with soil development.