Observation of BPTI and the dipeptide analog PaPb allows one to identify a number of residues that appear to be buried in the hydrophobic core of the protein fold. Comparing the contacted surface area of these residues (as calculated by the program dms) to the contacted surface areas of the isolated amino acids identifies the residues listed in the following table as core residues:
| Residue | Surface Area of Isolated AA | Surface Area of Residue in BPTI | Percent of SA buried in BPTI |
| Cys 5 | 35.61 sq. Å | 0.00 sq. Å | 100 % |
| Phe 22 | 65.97 sq. Å | 7.53 sq. Å | 89 % |
| Tyr 23 | 67.54 sq. Å | 2.22 sq. Å | 97 % |
| Cys 30 | 35.61 sq. Å | 0.20 sq. Å | 99 % |
| Phe 33 | 65.97 sq. Å | 4.75 sq. Å | 93 % |
| Phe 45 | 65.97 sq. Å | 9.11 sq. Å | 86 % |
| Cys 51 | 35.61 sq. Å | 0.00 sq. Å | 100 % |
| Cys 55 | 35.61 sq. Å | 0.00 sq. Å | 100 % |
Interestingly Asn 43 is also largely buried in the core, despite having a polar side group. Placing a hydrophilic group into a non-polar environment would be enthalpically costly, unless replacement H-bonds are available. In this case, Asn 43 makes specific hydrogen bonds to the backbone atoms of Tyr 23. (Also note that the hydroxyl of Tyr 23 protrudes from the core, rather than sacrifice external H-bonding opportunities.)
The small size of BPTI's hydrophobic core relative to its overall surface area is a problem. When BPTI folds, it buries the non-polar residues which yields an increased entropy for solvent molecules previously associated with clathrates surrounding those residues. However, this increase in solvent entropy is apparently insufficient to compensate the protein for its loss of chain entropy, and the overall DG is positive. By forming disulfide bonds between remote cysteines the peptide chain entropy of the unfolded protein can be reduced in exchange for the decrease in enthalphy associated with bond formation. The three disulfides appear to offer enthalpic compensation, in addition to the entropic benefits of the hydrophobic core, that make up for the loss of peptide conformational entropy.
When BPTI has reached a two disulfide intermediate (30-51/14-38 or 5-55/14-38) the last disulfide forms slowly. Study of the BPTI model shows that in either case, the remaining disulfide must be formed by residues that would be buried in the hydrophobic core. This environment restricts the conformational flexibility that the sulfhydryl groups need to reach the transition state for disulfide formation as well as hiding the groups from the necessary external oxidants.