This study analysed secondary school students' expressed and actual preferences for counsellor sex and race and also investigated the influence of type of problem on these preferences. A sample of 430 fifth form (15-16 years old) students including 74 Maori/Polynesian males, 109 Maori/Polynesian females, 90 pakeha (European) males, and 87 pakeha females from six schools in two cities completed the Choosing a Counsellor questionnaire. Half the sample were asked to assume they had a vocational problem while the remainder assumed they had a personal problem.Subjects were asked to first indicate how important 33 counsellor traits or characteristics would be to them when choosing a counsellor. 'Is the same sex as I am' and 'Is the same race as I am' were included in the list of items. Next, they were shown a slide/sound presentation of four school counsellors (one of each sex, one of each race) and asked to rank them in the order in which they would choose to see them. These two parts of the study made it possible to determine whether actual choice matched stated preference.The study confirmed the original hypotheses about students' preferences and choices regarding counsellor sex and race. Students did not express preferences for counsellor sex and race. However, their actual preferences were clearly otherwise: the majority chose a counsellor of the same sex and race as themselves. The influence of type of problem on these preferences and choices was minimal.In New Zealand secondary school guidance systems the staff are typically pakeha, which is the term used for Europeans by the indigenous people, the Maoris. However, the students seeking help from these staff reflect a much broader ethnic background. Less than three per cent of New Zealand's 275 secondary school counsellors are non-pakeha. In contrast, 14.3 per cent of all fifth form students are Maori (Department of Education, personal communication, March 1984) and this percentage does not include other minority groups which comprise at least a further 5 per cent of the general population. Similarly, while only 32.7