The resulting fire frequency of 2.9 × 10-4 per ship·year was based on a total of 93 incidents identified according to specific criteria. If these criteria had not been applied, the fire frequency would have been 2.6 × 10-3 per ship·year. None of the fire incidents contributing to the main result arose as a result of a preceding collision. The much lower rate of total losses from all causes for the UK compared with a world-wide basis is a good indication that the fire frequency quoted above is pessimistic. It is to beexpected that ships meeting the INF 2 or INF 3 requirements should suffer lower, or considerably lower, fire frequency in view of the additional fire safety features required by the INF Code, compared with those of general shipping from which the fire frequency data was derived.