Automatic evaluations are crucial for survival, but conscious self-reflection enables the formulation of nuanced evaluations to serve long-term goals. To operate effectively, both automatic and reflective evaluative processes need to integrate stored representations from previous experience (attitudes) with current contexts and goals, but contexts and goals have a more prominent role in reflective evaluation. Recent neuroimaging data provide new insights into the structure and function of evaluation and the dynamic ways that attitudes and reflective processing contribute to evaluation. In this paper, we propose a new iterative-reprocessing (IR) model of the neural bases of evaluation that highlights the role of the prefrontal cortex in the reprocessing of evaluative information. This model makes predictions that inform social-cognitive and cognitive-neuroscientific accounts of evaluation.