Examined the degree of reliability or agreement of assessments of dangerousness and compared decisions about dangerousness with decisions about patient management to determine how different professional groups from different countries compare in the evaluation of standardized information. 193 Ss from 6 countries--Brazil, Denmark, Egypt, Swaziland, Switzerland, and Thailand--and a variety of professional groups (e.g., magistrates, police officers, and non-medico-social professionals) considered 16 case histories. Ss were asked to indicate what degree of dangerousness they attributed to each individual and what they considered to be the optimal management. It was found that the level of agreement between Ss concerning the assessment of dangerousness was generally low: The level of 60% was reached for only 4 cases out of 16. Although it was expected that their specialized training and more objective rating behavior would lead them to present higher agreement as a group, psychiatrists did not reach a higher level of agreement on the ratings of dangerousness than nonpsychiatrists. Psychiatrists did show a significantly higher level of agreement on management decisions and also a tendency to rate individuals as more dangerous than did nonpsychiatrists. It is concluded that the results do not support the use of dangerousness as a scientifically or operationally valid concept.