During the past three years it has become possible to compute ab initi
o the C-13, N-15 and F-19 NMR chemical shifts of many sites in native
proteins. Chemical shifts are beginning to become a useful supplement
to more established methods of solution structure determination, and m
ay find utility in solid-state analysis as well. From C-13 NMR, inform
ation on phi, psi and chi torsions can be obtained, permitting both as
signment verification, and structure refinement and prediction. For N-
15, both torsional and hydrogen-bonding effects are important, while f
or F-19, chemical shifts are primarily indicators of the local charge
field. Chemical shift calculations are still slow, but shielding hyper
surfaces - the shift as a function of the dihedral angles that define
the molecular conformation - are becoming accessible. Over the next fe
w years, theoretical and computer hardware improvements will enable mo
re routine use of chemical shifts in structural studies, including the
study of metal-ligand interactions, the analysis of drug and substrat
e binding and catalysis, the study of folding/unfolding pathways, as w
ell as the characterization of conformational substates. Rather than s
imply being a necessary prerequisite for multidimensional NMR, chemica
l shifts and chemical shift non-equivalence due to folding are now beg
inning to be useful for structural characterization.