Abstract Most often, the unfolded state of peptides and proteins has been modeled as a statistical random coil. Here, we suggest an alternative model based on the presence of a significant, temperature-dependent conformational bias in the unfolded population. Conformational bias is suggested by our calculations [Proc. Natl. Acad. Sci. USA 96 (1999) 14258–14263], and it is found in recent studies of both proteins and peptides. The imposition of even a modest bias would transform our assessment of the folding problem.