A large number of studies demonstrated the importance of different HIV RNA structural elements at all stages of the viral life cycle. Nevertheless, the significance of many of these structures is unknown and plausibly new regions containing RNA structure-mediated regulatory signals remain to be identified. An important characteristic of genomic regions carrying functionally significant secondary structures is their mutational robustness, i.e. the extent to which a sequence remains constant in spite of mutations in terms of its underlying secondary structure. Structural robustness to mutations may prevent fitness loss due to disruption of a possibly functional conformation, pointing to the specific importance of the corresponding genomic region. In the current work we perform a genome wide computational analysis to detect signals of a direct evolutionary selection for mutational robustness in terms of secondary structure within HIV coding sequences. We believe that our analysis may enable a better understanding of viral evolutionary dynamics at the structural level and may benefit to practical efforts of engineering antiviral vaccines and novel therapeutic approaches.