The analgesic efficacy and safety of liposomal bupivacaine (LB) in third molar extraction was evaluated in this phase 3, double-blind, placebo-controlled study of subjects undergoing bilateral third molar extraction. Subjects were randomized 2 : 1 to infiltration with LB (133 mg/10 mL) or placebo, and received opioid rescue medication as needed. Primary efficacy measure was cumulative area under the curve (AUC) of numeric rating scale (NRS) pain severity scores through 48 hours (AUC of NRS_0–48) postsurgery. Other measures included AUC of NRS_0–24, AUC of NRS_0–72, and AUC of NRS_0–96, and incidence of adverse events. There were 150 subjects in the primary efficacy population (n = 99 LB, n = 51 placebo) and 89 in the per-protocol population (n = 59 LB, n = 30 placebo). Least-squares mean for AUC of NRS_0–48 was 172.3 LB versus 194.7 placebo (P = .227) in the primary efficacy population and 120.8 LB versus 183.3 placebo (P = .023) in the per-protocol population. At all time points, between-group differences in AUC of NRS scores were significant in the per-protocol population (LB lower than placebo, P < .05) but not in the primary efficacy population. The adverse event profile was similar between groups. LB produced significantly lower cumulative pain scores versus placebo at all time points in the per-protocol analysis but not in the primary efficacy analysis because of protocol violations. This study indicates significant improvement in pain scores in the third molar model, but because of extensive protocol violations additional studies are warranted to demonstrate effectiveness.