An ignition delay correlation has been developed for iso-octane based on the functional behavior exhibited by a detailed chemical kinetic mechanism. The correlation employs a traditional Arrhenius-based, power law formulation, tau = A phi(alpha)p(beta)chi o(2)(gamma) exp(lambda), including dependencies for equivalence ratio pressure (p) and oxygen percentage (chi o(2)). However the exponents for these parameters, alpha, beta, and gamma, respectively, are expressed as third-order polynomials with respect to temperature in order to capture changes in functionality seen across different regimes. At very low temperatures alpha, beta, and gamma are forced to a constant value, as seen within the mechanism. The activation energy term, lambda, is written as a combination of two quadratic expressions so that the behavior in the negative temperature coefficient (NTC) region can be captured. A pressure-dependent term is also included in the expression for; in order to reduce the activation energy at higher pressures in the NTC region due to increased low temperature reactivity, and the appearance of cool flame, or low temperature heat release (LTHR). The resulting expression contains 37 constants. The new correlation is applicable over a wide range of conditions and can be used for data comparisons and mechanism evaluation, as well as systems-level engineering simulations. in this work experimental data from rapid compression machines (RCM) and shock tubes (ST) are compared through normalizing features of the correlation, and the performance of a detailed kinetic mechanism is evaluated based on the functional behavior of the alpha, beta, gamma and lambda parameters. Six hundred and sixty-one (661) data points have been used to fit the 37 constants of the expression where the experimental conditions cover phi = 0.2-2.0, p = 1-60 atm, chi o(2) = 0.125-21% and T = 650-2000 K. Data normalized through the correlation indicate a standard deviation of 34%. Departures from the correlation can be attributed to an incomplete description of the functional dependencies, some inconsistencies with regard to diluent composition, experimental uncertainties, and facility-influenced phenomena. The experimental points have also been simulated using the LLNL detailed iso-octane mechanism, where the computed ignition delay times have been fit to the new correlation. Agreement and differences between the correlated experimental and simulation alpha, beta, gamma and lambda functions are highlighted, where it is noted that there is a substantial discrepancy concerning the increased low temperature chemistry at high pressure. implications for future engine design are discussed. (C) 2009 The Combustion Institute. Published by Elsevier Inc. All rights reserved.