This study analyzed the spatial dependencies of soil moisture and temperature in a six-stand chronosequence of boreal black spruce (Picea mariana (Mill.) BSP) stands. Spatial variability of soil temperature (T_SOIL) was evaluated twice during the growing season using four transects in each stand, employing a cyclic sampling design with measurements spaced 2–92 m apart. Soil moisture (θ_g) was measured on one occasion. A spherical model was used to analyze the geostatistical correlation structure; θ_g and T_SOIL at the 7- and 21-year-old stands did not exhibit stable ranges or sills. The fits with stable ranges and sills modeled the spatial patterns in the older stands reasonably well, although unexplained variability was high. Calculated ranges varied from 3 to 150 m for these stands, lengths probably related to structural characteristics influential in local-scale energy transfer. Transect-to-transect variability was significant and typically 5%–15% of the mean for T_SOIL and 10%–70% for θ_g. T_SOIL and θ_g were negatively correlated for most stands and depths, with T_SOIL dropping 0.5–0.9 °C for every 1% rise in θ_g. The results reported here provide initial data to assess the spatial variability of T_SOIL and θ_g in a variety of boreal forest stand ages.