Accident investigation programmes are presently used in many countries around the world, in an effort to reduce both the numbers of traffic accidents and injuries which result from those accidents. New Zealand has such a programme, which was started in 1985. The three organizations involved in the Accident Investigation Programme are: the Ministry of Transport, Transit New Zealand, and local authority government. With this combination of expertise, studies are conducted throughout the entire country. At the start of the programme, a systematic scheduling of studies to investigate the entire country during a five-year period was planned. Scheduling was revised in 1987 and the five year period will be completed in June 1992. The accident investigation programme will be continuing on from this period. Although other countries employ similar methods of accident investigation programmes, the systematic, country-wide approach in New Zealand is the first of its kind. In order to determine the effects of the programme, there must be a method of categorically recording the data obtained in the study. The accident investigation monitoring system was developed in 1989 for such a purpose. The accident investigation monitoring system database contains a combination of site characteristics and accident histories for each site studied under the programme. The scope for use of this data is large. The following thesis examines the accident investigation monitoring on an overall basis. New Zealand has a relatively small population, compared to other countries involved in accident investigation programmes. Because of the low numbers of accidents at sites in New Zealand, applications of well-known methods for calculating accident reduction do not produce conclusive results. Methods of reducing regression-to-mean and bias-by-selection, are tested, using the New Zealand data. Accident reduction calculated from the monitoring system data is used to determine the success of the programme.