Spatial variability of soil quality indicators was estimated in two studies to help plan sampling strategies. In the first study, spatial dependence of six soil quality indicators was measured for one soil order and one land use. Variation was measured between samples at 5‐, 30‐, and 100‐m spacings, and between fields in Allophanic soils under pasture, to estimate the effect of sample spacing on the spatial variability of pH, total volumetric C and total volumetric N, C:N ratio, Olsen P, and mineralisable N. The second study tested the effects of land use and soil type on variation between samples at spacings less than 29 m at 65 sites. This second study also included base saturation, cation exchange capacity, bulk density, total porosity, and macroporosity in addition to the indicators of the first study. All samples were taken from the soil surface to 10 cm depth. Results of the first study in Allophanic soils suggested sample replication spacings of 5 m or more for pH, 30 m or more for Olsen P, and mineralisable N, and 100 m or more for total C, total N, and C:N ratio. Sampling at less than these intervals will display some degree of autocorrelation and will less efficiently represent variation in the management unit being sampled. Results of the second study indicated no significant effect of soil order on coefficient of variation, but land use did have an effect, with cropping having the lowest coefficient of variation followed by pasture, exotic forest, and native forest.