OBJECTIVES: To identify the factors that influence the assessment of reported cases of physician-assisted death by members of the public prosecution. DESIGN/SETTING: At the beginning of 1996, during verbal interviews, 12 short case-descriptions were presented to a representative group of 47 members of the public prosecution in the Netherlands. RESULTS: Assessment varied considerably between respondents. Some respondents made more "lenient" assessments than others. Characteristics of the respondents, such as function, personal-life philosophy and age, were not related to the assessment. Case characteristics, i.e. the presence of an explicit request, life expectancy and the type of suffering, strongly influenced the assessment. Of these characteristics, the presence or absence of an explicit request was the most important determinant of the decision whether or not to hold an inquest. CONCLUSIONS: Although the presence of an explicit request, life expectancy and the type of suffering each influenced the assessment, each individual assessment was dependent on the assessor. The resulting danger of legal inequality and legal uncertainty, particularly in complicated cases, should be kept to a minimum by the introduction of some form of protocol and consultation in doubtful or boundary cases. The notification procedure already promotes a certain degree of uniformity in the prosecution policy.