Subleading effects in QCD global fits

I discuss several corrections to leading twist calculations of nucleon structure functions which are needed to include experimental data at large parton fractional momentum x and at low scales Q^2 in global fits of parton distribution functions. In particular I discuss the results of the CTEQ6X global fit, and some work in progress. Topics covered include the interplay of target mass and higher-twist corrections, the importance of nuclear corrections for deuterium target data, and applications to the study of quark-hadron duality. Implications for collider physics are highlighted.