Objective The aim of this study was to determine the differences in the efficacy and efficiency in providing critical care to hospitalized patients in teaching vs nonteaching intensive care units (ICUs) in Colombia. Methods A prospective cohort observational study was conducted. Location This study was conducted in 11 teaching and 8 nonteaching ICUs. From June 1 until December 31, 2005, data on 826 patients admitted consecutively to teaching ICUs and 825 patients admitted to nonteaching ICUs were analyzed. Measurements Acute Physiology and Chronic Health Evaluation II, Simplified Therapeutic Intervention Scoring System, ICU discharge status (dead or alive) and ICU length of stay, and standardized mortality ratios were considered in this study. A logistic regression and robust linear regression were performed. Results There were no differences in mortality (P = .25). Standardized mortality was less than 1 for both types of units. The teaching ICUs length of stay was 1 day longer (P < .01). Resource use is 25% higher in teaching units (P = .01). When the Simplified Therapeutic Intervention Scoring System score on the last day was from 21 to 35, a higher ratio of patients from the nonteaching ICUs was observed going floor or home when discharged from the ICU (P < .01). Conclusions Nonteaching ICUs discharge patients earlier than do teaching ICUs, but the effect of it remains to be clarified with further studies addressing questions as what happens after ICU discharge.