Disulfide bond formation experiment show that many factors strongly influence the stability changes, and that the stabilization by engineered disulfide bond is not always successful. The main reason is that the disulfide bond introduces a strain in the folded form that can reduce or reverse the stabilizing effect due to the reduction in the chain entropy of the unfolded form. The extent of destabilization of the folded form has now been directly measured for disulfide bonds introduced between two loops of Staphyloccocal nuclease. Destabilization ranges from 3.0-5.0 kcal/mole, which nearly exactly matches or dominates over the stabilization expected from the entropy reduction.

The other possible reason why a disulfide bond does not stabilizes the protein is that the cross-link induces formation of structure in the unfolded form, thereby stabilizing it. Betz (1992) have presented evidence of this for the 20-102 disulfide bond introduced in iso-1-cytochrome c.

Disulfide bond formation of iso-1-cytochrome c residue between 20 and 102