The relationship between affect and cognition has always been a key issue in psychology research, and over the past 50 years or so, there is a general “truth” in this field that cognitive consequences of positive and negative affect are generally insisted to be fixed. Immense amounts of empirical research revealed that positive affect is thought to trigger heuristic processing, a global focus and widened attention, while negative affect is thought to elicit systematic or elaborated processing, a local focus, and narrowed attention. Some relevant accounts have already been raised accordingly, and each of them has its explanation for this phenomenon. Then there has been a trend in latest literatures, however, a growing number of research demonstrates that the above relations could be reversed in some certain situation. When varying the psychological context, positive affect may also lead to detailed processing and a narrowed focus, and negative affect may lead to heuristic processing and a broadened focus. That is to say, not all research is consistent with this well-known “truth” and the relationship between affect and cognition is not fixed but flexible. Therefore, the most critical work about relationship between affect and cognition at present is how to explain these two opposite propositions. For the sake of regulating this paradox, a compatible account named Affect-as-Cognitive Feedback(AACF) has been invented by Isbell and his colleagues (2013) which is derived from the Affect-as-Information account(AAI). The core point of this account is that there is information contained in affect and the function of the information is to signal the value of accessibility thoughts or cognitive-processing style which was available in individuals’ minds. Positive affect confers positive value on accessible thoughts and processing inclinations, facilitating their use, while negative affect on the contrary confers negative value on such thoughts and inclinations, inhibiting their use. Thus, the AACF account implies that the impact of affect on cognition should be flexible in that its evaluative information depends on the momentary accessible mental content, and a fixed or dedicated connection between affect and cognitive-processing styles should be denied. We then reviewed evidence consistent with the view that a flexible impact of affect on cognitive-processing as a result of the relative accessible strategies. Such flexibility was shown for several outcomes for which evidence of a dedicated connection is usually assumed and seems robust, including attentional scope, impression formation, moral reasoning, and the implicit-explicit attitude correspondence. Compared with other accounts which support the flexible impact of positive and negative affect on cognitive-processing, AACF has its unique advantages on explanation effect and application range. Finally, we reviewed several expand directions for future studies combined with the AACF account. Subsequent researches should focus on the conceptual nature of affect and cognition to verify the necessity of this account, as well as the internal mechanism that why would consequence of positive and negative affect is respectively corresponding to adoption and abandonment of accessible strategies. The above exploration would shed light on affect-cognition research and make a better understanding of the affect-cognition relationship.