Models of tree growth responses to climate variability provide insight about the potential effects of global warming on forests. Using a unique dataset containing tree ringwidth measurements from all trees (>4 cm in diameter at breast height) in 1ha plots, we modelled time series of annual basal area increment (BAI) in subalpine forests of western Canada, which are expected to be highly sensitive to the effects of global warming. Our objective was to determine how BAI responses to inter-annual climate variable fluctuations varied among tree size classes, species, and sites. We tested multiple approaches for estimating climate effects on growth – which  incorporated a priori detrending of BAI for size, age, legacy, and competition effects – and settled on a parametric generalized least squares model.  We found that the climate signals in our time series of BAI were weak; < 6% of the interannual variance in growth was explained by climate variables. Nevertheless, there were clear patterns in growth-climate relationships related to tree size and species. BAI of the largest trees was most sensitive to climate variable fluctuations, and Callitropsis nootkatensis growth responses were significantly different in direction and magnitude than those of co-occurring Pinaceae species. Growth responses were idiosyncratic among sites.  While growth-climate relationships were stable over the 20th for large trees, the sensitivity of small trees to fluctuating climate variables decreased from the 1940s onward. Overall, our individual-tree and parametric modelling approaches emphasized the complexity of growth responses to climate variability and challenges associated with predicting future forest growth using tree rings.