Abstract
Following an initial investigation of the G.M. (brake light) conflict recording technique in various Canadian cities in 1972/73, a study was undertaken to define conflicts in such a way that a better link between these and accidents could be established. The conflict definition arising from this work was referred to as "post encroachment time" (PET) . In the new technique, six conflict types are recorded by teams of three or four observers, each of them being attributed a specific task; PET is timed for every event noted. Evaluation of the PET technique was carried out from 1978 to 1981, but due to small data sample size results were inconclusive.