Background: The current study examined correlates of cognitive insight in people enrolled in a methamphetamine pharmacotherapy trial; whether cognitive insight at the start of the trial predicted medication adherence and reductions in methamphetamine use during the trial; and, whether insight would remain stable over the trial or improve with reductions in methamphetamine use. Methods: A subset of people enrolled in a 12-week randomised placebo-controlled pharmacotherapy trial for methamphetamine dependence completed the Beck Cognitive Insight Scale, comprising subscales for Self-Reflection and Self-Certainty, at baseline (n = 152) and at week 12 (n = 79). Medication adherence was expressed as the percentage of non-missed doses measured using eCAP™ technology. Methamphetamine use days were assessed using the Timeline Followback. Results: At baseline, greater Self-Reflection was correlated with more severe methamphetamine withdrawal, and hostility, whereas Self-Certainty was correlated with less education and longer duration of methamphetamine use. No relationship was found between BCIS subscales at baseline and medication adherence (Self-Reflection b[SE] = −0.73 [0.43] p =.09; Self-Certainty b[SE] = −0.31 [0.48] p =.52,). Neither BCIS subscale was predictive of reduced methamphetamine use at 12 weeks (Self-Reflection b[SE] = 0.001 [0.01] p =.95 Self-Certainty b[SE] = −0.003 [0.01], p =.74). Self-Reflection decreased over the trial (t = 3.42, p =.001) but this was unrelated to change in methamphetamine use (Self-Reflection, b[SE] = −1.68 [1.16] p =.15) Change in methamphetamine use was found to be a significant predictor of Self-Certainty at 12 weeks (b [SE] = −2.71 [1.16] p =.02). Conclusions: We found no evidence that cognitive insight predicted medication adherence or methamphetamine reduction in people engaged in this trial. Ongoing or increased methamphetamine use predicted increased Self-Certainty at 12 weeks.