Abscissa, cocaine dose in mg/kg/inj (log-scale). Ordinate, total cocaine injections earned per day. All data points show the mean ± SEM in four monkeys. Repeated measures ANOVA found significant main effects of Treatment (F(3,18)=3.39; p<0.05) and Cocaine Dose (F(1,6)=45.40; p<0.001) as well as a significant Treatment X Cocaine Dose Interaction (F(3,18)=3.92; p<0.05). Dunnett’s post-hoc analysis showed that cocaine self-administration with lofexidine treatment was significantly lower than baseline when 0.01 and 0.032 mg/kg/inj cocaine was available (both p’s <0.01). Self-administration of each dose of cocaine was also compared to saline self-administration to determine which doses elicited significantly more behavior than when saline was available. Under baseline conditions, a significant effect of cocaine dose was found (F(4,12)=20.96; p<0.0001) and Dunnett’s post-hoc tests showed that 0.01 and 0.032 mg/kg/inj produced significantly more self-administration than saline. There was no main effect of cocaine dose with lofexidine treatment (F (4,12)=0.46; p>0.05). Dunnett’s tests indicate that cocaine self-administration did not differ from saline levels at any dose when lofexidine was administered as a pretreatment (all p’s >0.05). Paired t-tests on number of pellets earned found that responding for food was slightly, but significantly reduced during both food sessions (both p’s<0.05). * p<0.05 lofexidine treatment vs. cocaine baseline; † p<0.05 cocaine vs. saline self-administration.