Some evidence suggests that motivational approaches are less efficacious – or even counter-productive – with persons who are relatively motivated at baseline. The present study was conducted to examine whether disordinal moderation by baseline motivation could partially explain negative findings in a previous study [Winhusen, T., Kropp, F., Babcock, D., Hague, D., Erickson, S.J., Renz, C., Rau, L., Lewis, D., Leimberger, J., Somoza, E., 2008. Motivational enhancement therapy to improve treatment utilization and outcome in pregnant substance users. J. Subst. Abuse Treat. 35, 161–173]. Analyses also focused on the relative utility of the University of Rhode Island Change Assessment (URICA) scale, vs. a single goal question as potential moderators of Motivation Enhancement Therapy (MET). Participants were 200 pregnant women presenting for substance abuse treatment at one of four sites. Women were randomly assigned to either a three-session MET condition or treatment as usual (TAU). Generalized Estimating Equations (GEE) revealed no significant moderation effects on drug use at post-treatment. At follow-up, contrary to expectations, participants who had not set a clear quit goal at baseline were less likely to be drug-free if randomized to MET (OR = 0.48); participants who did set a clear quit goal were more likely to be drug-free if randomized to MET (OR = 2.53). No moderating effects were identified via the URICA. Disordinal moderation of MET efficacy by baseline motivation may have contributed somewhat to the negative results of the [Winhusen, T., Kropp, F., Babcock, D., Hague, D., Erickson, S.J., Renz, C., Rau, L., Lewis, D., Leimberger, J., Somoza, E., 2008. Motivational enhancement therapy to improve treatment utilization and outcome in pregnant substance users. J. Subst. Abuse Treat. 35, 161–173] study, but in the opposite direction expected. A simple question regarding intent to quit may be useful in identifying persons who may differentially respond to motivational interventions. However, moderation effects are unstable, may be best identified with alternate methodologies, and may operate differently among pregnant women.