This study assessed the variability of sample estimates for downed and dead woody fuel weight in natural lodgepole pine (Pinuscontorta Dougl.) stands using line-intersect sampling procedures. Equilateral triangles (30 m/side) were established at each of 40 sample sites with variable length transects on each side to estimate fuel weights by diameter class. Regardless of the number of sides measured, the standard error for fuels less than 7.0 cm was at most 20% of the mean. Even measuring only one side of the triangle, using a single transect instead of the triangular sample unit, still achieved standard errors less than 20% of the mean. Standard errors for classes greater than 7.0 cm were all greater than 20% of the mean. For these classes, more samples are required to achieve the 20% standard error limit; however, depending on costs, the triangular sample unit may not be the best solution. In this study, intracluster correlations were above 0.7 for the fuel diameter classes greater than 7.0 cm, suggesting that multiple transects at a given sample location contribute little new information. This effect, although less pronounced, was also observed with the smaller diameter classes.