A human laboratory model of intravenous methamphetamine self-administration may facilitate study of putative treatments for methamphetamine addiction. We conducted a double-blind, placebo-controlled, between groups investigation of the acetylcholinesterase (AChE) inhibitor rivastigmine in non-treatment-seeking volunteers who met criteria for methamphetamine abuse or dependence. Safety and subjective effects data derived from days 1-10 of this protocol are described in a separate publication. In this report, we describe self-administration outcomes in participants randomized to treatment with rivastigmine (0 mg, N=7; 1.5 mg, N=6; 3 mg, N=9); data that were collected on days 11-15 of the inpatient protocol. On day 11, participants sampled two infusions of methamphetamine (0 and 30 mg, i.v.). On days 12-15, participants made ten choices each day to receive an infusion of either methamphetamine (3 mg, IV) or saline or a monetary alternative ($0.05-$16). The study design allowed for evaluation of differences in behavior on days in which infusions were performed by the physician (experimenter-administered) versus by the participant using a PCA pump (self-administered), and when monetary alternatives were presented in either ascending or descending sequence. The data show that rivastigmine (1.5 and 3 mg), as compared to placebo, did not significantly alter total choices for methamphetamine (p=0.150). Importantly, the number of infusion choices was greater when methamphetamine was available then when saline was available (p<0.0001), and the number of money choices was greater when saline was available then when methamphetamine was available (p<0.0001). The total number of choices for methamphetamine was not altered as a function of a participant's preferred route of methamphetamine use (p=0.57), and did not differ significantly whether they were experimenter-administered or self-administered (p=0.30). In addition, total choices for methamphetamine were similar made when money was available in an ascending versus descending sequence (p=0.49). The participants' years of methamphetamine use, recent use of methamphetamine (in the past 30 days), or baseline craving (indexed here as "Desire") on the day of the self-administration task were not predictive of number of choices for methamphetamine. In a subset of participants (N=8) for which data was available, individual dose of methamphetamine (3 x 3 mg, i.v.) produced significant increases in positive subjective effects, and a preliminary analysis revealed that 3 mg rivastigmine was associated with reductions in these responses, as compared to placebo. In summary, the current report indicates that there were no effects of rivastigmine on total choices for methamphetamine, that there were low levels of methamphetamine self-administration but these were 8 times greater than saline, and that choice behavior was insensitive to alternative reinforcers. In addition, we showed that rivastigmine may reduce the positive subjective effects produced by methamphetamine during self-administration.