Background & Aims: Hospitals are the most important centers that provide health services in the country. They have to use scientific and efficient management at all levels for promoting health. Intensive care units (ICU) are one of the most sensitive parts of the hospitals; therefore, paying attention to its efficiency is of particular importance. The purpose of this paper was to estimate the technical efficiency of intensive care units in teaching hospitals in Kerman, Iran. Methods: This descriptive-analytic study was done in 2009 in Kerman teaching hospitals. Input variables included the number of staff (physicians, nurses and other staff), number of equipments, number of beds and output variables include the percentage of bed occupancy. Data were collected by checklists and forms built by the researchers and were analyzed by applying stochastic frontier analysis (SFA) by the Frontier 4.1 software. Results: The results of stochastic frontier analysis in intensive care unit showed that the capacity to improve efficiency of intensive care units in these hospitals could increase to 21% (compared with the most efficient hospitals in the studied population). The ICU No. 3 of Bahonar Hospital had the lowest technical efficiency (0.414) in April and maximum technical efficiency was of Afzalipour Hospital intensive care unit in January (0.951); Moreover, given minor elasticity coefficients indicated that their sum was more than one, and as a result, it showed a rising in productivity scale. Conclusion: Results of analysis of surplus and an additional use of inputs produced in intensive care units suggested that efficiency achieved in August was more than the other months; it is noteworthy that the average of technical efficiency in these sectors has not have a stable process.