Experimental kinetic data on the reactions of pure chemicals, destruction of heavy hydrocarbons, and gas formation have been combined with thermodynamic theory and empirical data on oil and gas occurrences to constrain the range of plausible activation energies and frequency factors for oil destruction and gas formation in nature. It is assumed explicitly here that the kinetics of oil destruction and gas formation can be adequately described using a set of parallel first-order reactions. At geologic temperatures and pressures the mean activation energy for oil destruction and gas generation is about 59 kcal/mol (246.9 kJ/mol), with a frequency factor of about 10(14.25) s(-1) (1.78.10(14) s(-1)). A narrow distribution of activation energies [sigma=1.5 kcal/mol (6.3 kJ/mol)] for destruction of oil seems intuitively more reasonable than a single activation energy, and also seems to fit empirical data on high-temperature occurrences of condensate slightly better. No large or systematic variation in cracking rates or kinetics is apparent for different oil types. Using these recommended kinetic parameters, the maximum temperature at which oil can be preserved as a separate phase varies from about 170 degrees C at geologically very slow heating rates to slightly over 200 degrees C at geologically extremely fast heating rates. Using this model, oil destruction occurs at slightly higher temperatures than those predicted by older kinetic models, but at considerably lower temperatures than those suggested by some recent studies. Differences in predicted levels of cracking obtained from the various models in use today can affect exploration decisions. (C) 2000 Elsevier Science Ltd. All rights reserved.