Oil and gas accident statistics reveal that workers' potential for injury or death from occupational accidents is at least as high as that associated with explosions, fires, and other major incidents. The authors' contribution to the ongoing efforts to improve the situation is a holistic, quantitative model capable of predicting occupational accident frequency. Model inputs include external, corporate and direct factors, and the approach includes methods often favoured by professionals from the safety, engineering, and psychology disciplines. As a precursor to model development, a comprehensive literature review was conducted. One objective of the review was to understand the previous approaches taken by other researchers and thereby identify any gaps in the knowledge. The lack of a holistic, quantitative model addressing occupational accidents in the oil and gas industry was established. A second objective was to select and categorise the factors most influential in the accident process and thereby provide a foundation for the present model. Optimal choices required a systematic study of the factors proposed by other researchers. The literature studied has been summarised in this paper, subdivided into two primary groups: (i) descriptions of existing accident models, and (ii) analyses of existing data. In addition, descriptions of literature specifically concerned with two key elements in the occupational accident process, human factors and safety culture, have been included. Based on the literature reviewed, the novelty of the present model has been discussed. The systematic approach taken to choose the factors thought most influential in the occupational accident process, and which were therefore included in the model, has been described. Some details of the proposed model have been included. (C) 2006 Elsevier Ltd. All rights reserved.