This experiment explored how mock-jurors’ (N = 648) guilt decisions, perceptions of the defendant, memories, and evidence interpretation varied as a function of jury type and pretrial publicity (PTP); utilizing a 2 (jury type: pure-PTP vs. mixed-PTP) × 3 (PTP: defendant, victim, and irrelevant) factorial design. Mock-juries (N = 126) were composed of jurors exposed to the same type of PTP (pure-PTP; e.g., defendant-PTP) or different types of PTP (mixed-PTP; e.g., half exposed to defendant-PTP and half to irrelevant-PTP). Before deliberations jurors exposed to defendant-PTP were most likely to vote guilty; while those exposed to victim-PTP were least likely. After deliberations, jury type and PTP affected jurors’ guilt decisions. Specifically, jurors deliberating on pure-PTP juries had verdict distributions that closely resembled the predeliberation distributions. The verdict distributions of jurors on mixed-PTP juries suggested that jurors were influenced by those they deliberated with. Jurors not exposed to PTP appeared to incorporate bias from PTP-exposed jurors. Only PTP had significant effects on postdeliberation measures of memory and evidence interpretation. Mediation analyses revealed that evidence interpretation and defendant credibility assessments mediated the effect of PTP on guilt ratings. Taken together these findings suggest that during deliberations PTP bias can spread to jurors not previously exposed to PTP. In addition, juries composed of jurors exposed to different PTP slants, as opposed to a single PTP slant, can result in less biased decisions. Finally, deliberating with others who do not share similar biases may have little, if any, impact on biased evidence interpretation or memory errors.