Facial expressions serve as social cues that shape decision processes in social interactions. This study investigated how distinct functional smiles shape fairness decisions by combining behavioral, model-based (HDDM), and ERP analyses. Thirty participants acted as responders in an ultimatum game where proposers displayed enjoyment, affiliative, or dominance smiles preceding fair or unfair offers. Behaviorally, a significant emotion × offer interaction emerged: participants rejected unfair offers far more often than fair ones, demonstrating a robust fairness preference. Under unfair conditions, however, enjoyment smiles significantly reduced rejection compared with all other emotions, indicating that positive expressions mitigated participants’ aversive reactions to inequity. Model-based analyses (HDDM) of the unfair condition revealed drift rates (v) closest to zero for enjoyment smiles, reflecting slower evidence accumulation when cooperative intent conflicted with inequitable treatment. At the neural level, early N400 amplitudes showed main effects of offer type and emotion, being smallest for fair offers and enjoyment smiles—conditions most consistent with expectations of fairness and cooperation. The CPP captured the integration of emotional and fairness information, showing an emotion × offer interaction driven by unfair offers; under these conditions, enjoyment smiles elicited larger late amplitudes before response, indicating that when evidence accumulation is slower, CPP activity intensifies during the final decision stage. Together, these findings suggest that positive emotional cues recalibrate fairness decisions by modulating the temporal dynamics of evidence integration, illustrating how positive signals from the proposer can temper—but not overturn—people’s aversion to unfairness.