An alternative procedure to predict both high-temperature stage and cool flames ignition delays under transient thermodynamic conditions is intended to be validated in this paper. An experimental study has been carried out in a Rapid Compression-Expansion Machine (RCEM), using different iso-octane/n-heptane blends in order to cover a wide range of octane numbers (from 25 to 75) under a wide range of initial temperatures (from 363 K to 423 K), compression ratios (14 and 19), 02 molar rates (from 21% to 16%) and equivalence ratios (from 0.4 to 0.8). The results obtained have been used to validate direct chemical kinetic simulations, as well as to evaluate the alternative predictive method and the Livengood & Wu integral method. Simulations have been performed solving a detailed chemical kinetic mechanism in CHEMKIN. The experimental results show good agreement with the chemical kinetic simulations and with the alternative predictive method. In fact, the mean relative deviation between experiments and simulations is equal to 1.7%, 2.2% and 3.1% for PRF25, PRF50 and PRF75, respectively. Besides, the alternative method has shown good predictive capability not only for the high-temperature stage of the process, but also for cool flames, being the mean relative deviation versus the experimental data lower than 3.3% for all fuels. Better predictions of the ignition delay have been obtained with the alternative procedure than the ones obtained with the classic Livengood & Wu expression, especially in those cases showing a two-stage ignition pattern, in which the Livengood & Wu integral method is not able to predict the high-temperature stage of the process. (C) 2016 Elsevier Ltd. All rights reserved.