Potential uses of dry weight information for simulation, and sources of data, for simulation of forest fuels were discussed. Tree and crown component dry weights of 23 Douglas fir, 18 western hemlock, and 23 western red cedar trees expanded to 314 "model" trees were investigated. Data were collected on the University Research Forest near Haney, British Columbia. Multiple regression techniques were used for the analysis. The best four independent variables are listed and the possibility of retaining a single independent variable, with its implications, are discussed. All predicting equations were conditioned to zero intercept and statistically tested for significance by the F-test. Of the independent variables tested, the combined variable of breast height diameter squared times total tree height was most closely related to tree component dry weights. The product of diameter and crown width squared was found to be the best single variable to describe crown component dry weights for Douglas fir. The dry weights of western hemlock and western red cedar crown components were most closely associated with breast height diameter times crown length. Reliable estimates of tree component dry weights were obtained. Crown components were highly variable and widely dispersed about the mean which resulted in less accurate estimates. / Forestry, Faculty of / Graduate