Eighty-four mature Norway spruce (Picea abies L. Karst), silver fir (Abies alba Mill) and Scots pine (Pinus sylvestris L.) trees were winched over to determine the maximum resistive turning moment (M(a)) of the root-soil system, the root-soil plate geometry, the azimuthal orientation of root growth, and the occurrence of root rot. The calculation of M(a), based on digital image tracking of stem deflection, accounted not only for the force application and its changing geometry, but also for the weight of the overhanging tree, representing up to 42% of M(a). Root rot reduced M(a) significantly and was detected in 25% of the Norway spruce and 5% of the silver fir trees. Excluding trees with root rot, differences in M(a) between species were small and insignificant. About 75% of the variance in M(a) could be explained by one of the four variables--tree mass, stem mass, stem diameter at breast height squared times tree height, and stem diameter at breast height squared. Among the seven allometric variables assessed above ground, stem diameter at breast height best described the root-soil plate dimensions, but the correlations were weak and the differences between species were insignificant. The shape of the root-soil plate was well described by a depth-dependent taper model with an elliptical cross section. Roots displayed a preferred azimuthal orientation of growth in the axis of prevailing winds, and the direction of frequent weak winds matched the orientation of growth better than that of rare strong winds. The lack of difference in anchorage parameters among species probably reflects the similar belowground growth conditions of the mature trees.