Deciding whether a patient with a disorder of consciousness -DOC- is in an unresponsive aware syndrome or a minimal conscious state is challenging. Lack of sensitive and objective diagnostic techniques determines that about 40% of cases are misdiagnosed. The situation is more complicated in intensive care environment where a wrong diagnosis could affect the quality of treatment. In this study as part of a protocol that aim to standardize the cognitive assessment of critically ill patients we employed two different stimulation designs: 1- a MMN/P300 ERP paradigm to the patient's name uttered by a familiar -ONF- and unfamiliar voice -ONUF; 2- different structured and non-structured auditory conditions (ex: noise, music, message) while recording EEG. To better characterized patients with and without DOC we combined time-frequency, source localization and EEG functional connectivity analyzes. The results showed that ONF and ONUF in conscious patients and in a patient with locked-in-syndrome elicited similar event-related potentials with a MMN and a P300 components that were characterized by a synchronization in theta range between 300-700 ms in the time-frequency analysis. IS modeled with LORETA evinced neural generators with maxima in bilateral anterior frontal regions as well as in parietoccipital and temporal areas. However patients in minimal conscious state and in unresponsive aware syndrome differed in the absence of theta synchronization and of bilateral frontal and parieto-occipital involvement. Main differences in the EEG of conscious and unconscious patients obtained during auditory stimulation were: 1- distinctive activations of frontoparietal areas, including precuneus and cuneus in conscious patients, especially during the presentation of structured stimuli (ie: music, message); 2- A pattern of synchronization at the level of electrodes characterized by an increase in the clustering and in the strength of connections during the presentation of structured stimuli in conscious patients. This study showed the feasibility to incorporate different stimulation strategies and post-processing techniques as routine tools in the evaluation of patients with DOC. Results suggested that the combination of these analysis techniques could have enough sensitivity to and could enhance the likelihood to detect residual -preserved- cognitive processing in difficult cases with altered consciousness state in ICU.