In this talk, I will present work on the neural and cognitive mechanisms underlying value-based de-cisions, with a focus on multi-attribute choices. Starting with findings from eye-tracking experi-ments, I will discuss how individuals allocate visual attention in a goal-directed manner to efficiently identify the most suitable option, as captured by a hierarchical Bayesian model of information search. The talk then shifts to a novel EEG study testing the neural mechanisms along the predic-tions of our theory, the Multi-Attribute Search and Choice (MASC) model, which posits that belief updating during fixations guides decision processes. EEG analyses reveal neural signatures—a P3-like component and centro-parietal positivity—that correlate with model-derived belief updates, providing neural evidence for Bayesian inference in active decision-making. Together, these studies demonstrate how attention and belief updating dynamically shape complex choices.