Recently, Hasserodt et al. proposed new HIV-1 drug candidates based on a weak N center dot center dot center dot CO interaction, designed to be a close transition state analog (Gautier et al. Bioorg. Med. Chem. 2006, 14, 3835-3847; Waibel et al. J. Bioorg. Med. Chem. 2009, 17, 3671-3679). They suggested that further improvement of these compounds could take advantage of computational approaches. In the present work, we propose an atomistic model based on a QM/MM description of the N center dot center dot center dot CO core embedded in an amino-aldehyde peptidic inhibitor. We focus on the existence of the N center dot center dot center dot CO interaction in the aqueous and enzymatic media. We show that the N center dot center dot center dot CO bond holds in water, while in the protein, there is a competition between the formation of the weak N center dot center dot center dot CO bond and the conservation of the hydrogen bond network around the structural water molecule W301 that is known to be crucial for the binding of bath substrates and inhibitors. This competition hampers the inhibitor to provide strong stabilizing interactions with all the key parts of the protein at the same time. Our calculations indicate that this competition we observed in peptidic compounds might be avoided by the proper design of nonpeptidic ones, following a similar strategy to that for cyclic urea derivatives and the FDA approved drug Tipranavir. Hence, our results encourage further development of the nonpeptidic hydrazino-urea derivatives suggested recently by Hasserodt et al.