Tree growth in sub-humid forests is influenced by environmental conditions that may differ across local topographic gradients. My study investigated how the growth of two conifer species (Pinus contorta var. latifolia and Picea glauca) varied across a range of habitats in a water-limited island forest landscape in western Canada. Historical precipitation from 1951-2016 had a positive overall effect on annual radial growth (as measured by tree rings), but the magnitude of this effect varied with proximity to a permanent water source and tree size. Mean annual precipitation had a greater positive influence on the growth of larger trees and on those growing far from water sources, a weaker positive effect on both species at intermediate distances from water, and the weakest influence on P. glauca trees growing close to water sources. These results indicate that tree growth rates may change if there is greater inter-annual variability of precipitation under climate change, though the magnitude and direction of these responses could be quite variable. The processes involved in tree growth and development can affect seasonal patterns of radial growth and wood density, which change across both space and time. I statistically partitioned the variation in four separate measures of growth (earlywood width, earlywood density, latewood width, and latewood density) across years, plots, and trees. Both ring width and wood density were more variable among individuals within sampling plots than they were among plots or between years. Most pairs of growth variables were positively correlated across years and individuals in the two species, but correlations between growth variables were mostly uncertain across sampling plots. An exploratory analysis revealed a number of tentative associations with environmental variables: temperature had a negative relationship with growth and latewood density; precipitation had a negative relationship with earlywood density; P. glauca growth was highest in aspen-dominated stands; P. contorta growth was negatively related to plot basal area; height and diameter had positive relationships with growth; age and slenderness had negative relationships with growth. A better understanding of variation in tree growth and wood density may lead to new insights into how wood production responds to environmental factors in a changing climate.