Conversely to behaviorist ideas, recent studies suggest that introspection can be accurate and reliable. However, an unresolved question is whether people are able to report specific aspects of their phenomenal experience, or whether they report more general nonspecific experiences. To address this question, we investigated the sensitivity and validity of our introspection for different types of conflict. Taking advantage of the congruency sequence effect, we dissociated response conflict while keeping visual conflict unchanged in a Stroop and in a priming task. Participants were subsequently asked to report on either their experience of urge to err or on their feeling of visual conflict. Depending on the focus of the introspection, subjective reports specifically followed either the response conflict or the visual conflict. These results demonstrate that our introspective reports can be sensitive and that we are able to dissociate specific aspects of our phenomenal experiences in a valid manner.