The relation between 28-day strength of normal cured concrete and accelerated strength is investigated by using an ordinary portland cement and a trass cement under two different accelerated curing conditions, warm water and boiling water. Linear regression analysis was applied on the test results and evaluated by using the efficiency concept, i.e., the ratio of accelerated strength and 28-day normal cured strength. It is concluded that the ordinary portland cement gives higher efficiency than that of the trass cement. The difference due to the cement type is less in the boiling water method than that in the warm water method. (C) 2001 Elsevier Science Ltd. All rights reserved.