Although constraint relaxation had been proposed by Knoblich and colleagues in 1990s as a key process in insight problem solving, the involved procedure and stages of information processing and the associated brain mechanisms have not yet been investigated in detail. From a theoretical perspective, the present paper for the first time proposed a preliminary account of the processing stages in hinted constraint relaxation, based on evidence from a logogriph insight problem solving task involving the use of hinted answers. According to this account, in insight problem solving, hinted constraint relaxation consists of three different information processing stages. The first stage involves pre-warning from conflicting messages or preprocessing operations characterized by the engagement of early attention. The second stage is the crucial process of constraint relaxation characterized by the alternation between the old and new ways of thinking. As this process mainly involves the expansion of the space for basic problem representation, the right hemisphere is suggested to play a dominant role during the process. The third stage is the process of re-integration characterized by top-down controlled processing. In support of this account, the ERP results showed that the first stage manifested itself as the N100/P200 complex accompanying constraint relaxation in insight problem solving;that the second stage manifested itself as the P300 varying in the 300-400ms time window illustrated with scalp topography and difference waves;and that the third stage manifested itself as changes in N400 in the 400-800ms time window. Furthermore, the ERP results showed that in stage 2, or the crucial stage for constraint relaxation, the right brain showed significantly stronger activity than the left brain, suggesting that the expansion of the basic problem representation space relies more on the right brain. In contrast, an opposite trend in the pattern of hemisphere lateralization was found in stage 3 or the stage for information re-integration. These findings offer a new view in understanding the cognitive and neural mechanisms of constraint relaxation in insight problem solving.