Objectives: African American youth who use marijuana are less likely to attend and complete treatment than white youth. Limited information is available on racial and age variation in treatment attendance and completion among adults who use marijuana. Methods: The current research examined differences in communitybased substance abuse treatment attendance and completion between adult African American and white marijuana users in 2 independent samples from the US southeastern (N = 160; 70.6% African American) and mid-Atlantic (N = 450; 34.7% African American) regions. Outcomes: Attended at least 3 treatment sessions, successful treatment completion, number of days in treatment, and percentage of positive urine drug screens. Adjusted regression models examined the association of race, age, and the interaction of race and age with treatment attendance and completion. Results: In the southeastern sample, successful treatment completion was significantly associated with the interaction of race and age (adjusted odds ratio = 1.35, 95% confidence interval = 1.08-1.69); whereas younger African Americans were less likely to complete treatment than older African Americans, age was unrelated to treatment completion among whites. In the mid-Atlantic sample, African Americans were significantly less likely to attend at least 3 treatment sessions (adjusted odds ratio=0.37, 95% confidence interval= 0.23-0.58), and younger adultmarijuana users were retained for fewer days in treatment (adjusted β = 0.13, 95% confidence interval = 0.27-2.48). Among African Americans, 37.9% (SD = 38.0) of urine drug screens tested positive for at least 1 illicit drug, and among whites, 34.2% (SD = 37.8%) tested positive; the percentage of positive urine drug screens was not associated with race or age. Conclusions: Among marijuana-using adults, treatment attendance and completion differ by race and age, and improvements in treatment completion may occur as some African Americans mature out ofyoung adulthood. © 2014 American Society of Addiction Medicine.