Utilizing the recently developed ONIOM-ccCA methodology, the proton affinities (PAs) of small biomolecules (∼15 nonhydrogen atoms) were determined. The ONIOM-ccCA method was used to predict the PAs of eight amine-containing molecules and these predictions were then compared with experiment. In these comparisons, a protocol for the determination of the model system size was established, and the low level method and basis set for the real system component of ONIOM-ccCA were considered. Several possible density functionals were investigated for description of the real system (low layer), including B3LYP, B97-1, B97-2, B98, BMK, M06, and M06-2X. The resulting proton affinities were compared both to experiment and to theoretical values that were calculated using the correlation consistent Composite Approach (ccCA). The impact of the choice of augmented correlation consistent basis sets in the ONIOM-ccCA low level calculation was also examined, and PAs calculated with augmented and non-augmented basis sets were compared to those extrapolated to the complete basis set (CBS) limit. ONIOM(ccCA:B3LYP/aug-cc-pVDZ) was then applied to deoxyribonucleosides to determine the PAs at possible sites of protonation.