【正文】 n, monitoring and prognosis of accident developments, the detection of positive and negative safety developments, the definition of safety targets and the(product)evaluation of long term and large scale safety application of accident analysis is strongly limited for problem analysis, prospective and retrospective safety analysis on newly developed traffic systems or safety measures, as well as for(process)evaluation of special short term and small scale safety is an urgent need for the analysis of accidents in real time, in bination with background behavioural incident detection, bined with video recording of accidents may soon result in financially acceptable type of research may eventually lead to a better understanding of the concept of risk in traffic and to wellestablished paper is primarily based on personal experience concerning traffic safety, safety research and the role of accidents analysis in this experiences resulted in rather philosophical opinions as well as more practical viewpoints on research methodology and statistical number of these findings are published already this lack of direct observation of accidents, a number of methodological problems arise, leading to continuous discussions about the interpretation of findings that cannot be tested a fruitful discussion of these methodological problems it is very informative to look at a real accident on then turns out that most of the relevant information used to explain the accident will be missing in the accident studies also cannot recollect all the data that is necessary in order to test hypotheses about the occurrence of the a particular carcar accident, that was recorded on video at an urban intersection in the Netherlands, between a car ing from a minor road, colliding with a car on the major road, the following questions could be asked:Why did the driver of the car ing from the minor road, suddenly accelerate after ing almost to a stop and hit the side of the car from the left at the main road? Why was the approaching car not noticed? Was it because the driver was preoccupied with the two cars ing from the right and the gap before them that offered him the possibility to cross? Did he look left before, but was his view possibly blocked by the green van parked at the corner? Certainly the traffic situation was not the moment of the accident there were no 5bicyclists or pedestrians present to distract his attention at the regularly overcrowded parked green van disappeared within five minutes, the two other cars that may have been important left without a is hardly possible to observe traffic behaviour under the most relevant condition of an accident occurring, because accidents are very rare events, given the large number of the new video equipment and the recent developments in automatic incident and accident detection, it bees more and more realistic to collect such data at not too high to this type of data that is most essential for a good understanding of the risk increasing factors in traffic, it also important to look at normal traffic behaviour as a reference question about the possibilities and limitations of accident analysis is not lightly cannot speak unambiguously about accident analysis covers a whole range of activities, each originating from a different background and based on different sources of information: national data banks, additional information from other sources, specially collected accident data, behavioural background data answer the question about the possibilities and limitations, we first have to look at the cycle of activities in the area of traffic of these activities are mainly concerned with the safety management of the traffic system, some others are primarily research following steps should be distinguished:description of the problem and its main characteristics。selection and implementation of safety measures。the probability of an accident to occur is independent from the occurrence of previous accidents。the occurrence of accidents is homogeneous in these two assumptions hold, then accidents are Poisson first assumption does not meet much are rare events and therefore not easily influenced by previous some cases where there is a direct causal chain(., when a number of cars run into each other)the series of accidents may be regarded as one plicated accident with many cars assumption does not apply to are often related to the same accident and therefore the independency assumption does not second assumption seems less obvious at first occurrence of accidents through time or on different locations are not equally , the assumption need not hold over long time is a rather theoretical assumption in its it holds for short periods of time, then it also holds for long periods, because the sum of Poisson distributed variables, even if their Poisson rates are different, is also Poisson Poisson rate for the sum of these periods is then equal to the sum of the Poisson rates for these assumption that really counts for a parison of(posite)situations, is whether two outes from an aggregation of situations in time and/or space, have a parable mix of basic ., the parison of the number of accidents on one particular day of the year, as pared to another day(the next day, or the same day of the next week etc.).If the conditions are assumed to be the same(same duration, same mix of traffic and situations, same weather conditions etc.)then the resulting numbers of accidents are the outes of the same Poisson assumption can be tested by estimating the rate parameter on the basis of the two observed values(the estimate being the average of the two values).Probability theory can be used to pute the likelihood of the equality assumption, given the two observations and their statistical procedure is rather Poisson assumption is investigated many times and turns out to be supported by a vast body of empirical has been applied in numerous situations to find out whether differences in observed numbers of