The Young's modulus of cubes of human cancellous bone was measured in three orthogonal directions. Apparent density and mineral volume fraction were also measured, as were two architectural variables, fabric and connectivity, which were determined using image analysis techniques. Multiple regression was used to relate the Young's modulus to the four explanatory variables. The results from this study are compared with those obtained from a previous investigation using non-human cancellous bone. The relationships revealed by the two studies are very similar. It was possible to explain approximately 93 per cent of the variance in Young's modulus using the four variables in this present study. Apparent density is the major explanatory variable in both studies and shows a strong correlation with connectivity. In common with the non-human study the measure of fabric is a worthwhile explanatory variable; however, connectivity and mineral volume fraction are relatively unimportant. The four explanatory variables contribute to a successful model for the prediction of Young's modulus. Any other candidate variables are likely to be unimportant or be highly correlated with those already investigated.