The aim of the study was to explore treatment efficiency in an undergraduate comprehensive dental care program (CDCP). The study sample consisted of the records of 652 patients from the CDCP of the School of Dentistry, Federal University of Goias, Brazil, who were treated in the period from 2004 through 2009. A total of 45 clinical procedures performed by the students was listed and a panel of 19 judges, graded the perceived complexity of each procedure on a 11-point scale using an adaptation of the Thurstone method. Spearman's correlation, one-way Anova, Kaplan-Meier, and Cox regression were used to build a predictive model for time-to-event data—completion of treatment (CT). Treatment time for CT was correlated with complexity scores (; ). The average estimated median months for CT was 23.0 (95%CI = 19.6–26.3) and was significantly different () among complexity levels (low 13.0, intermediary 19.0, high 47.0). When low complexity was the reference category, estimated changes in risk for incomplete treatment were greater for intermediary (; 95%CI = 0.40–0.75) and high complexity cases (; 95%CI = 0.23–0.45). The results indicated that treatment complexity has a large influence on undergraduate CDCP efficiency and should be considered when planning organizational strategies for the clinical environment.