The reliability, content validity, and construct validity were compared for two test formats in a public examination used to assess a secondary school geography course. The 11-item geography portion of the Intermediate Certificate Examination (essay examination) was administered in June 1987 to 400 secondary school students in Ireland who also took an objective 43-item multiple-choice geography test about 7 months later (in November 1987). The scores on the essay test were less reliable and were susceptible to a significantly larger standard error of measurement than were those on the objective test. Test format was also found to have an effect on classroom practices. The two tests measured different skills, and there was insufficient evidence to prove that writing ability or quality of handwriting influenced scores assigned to essays. It is suggested that a mixed test format is a desirable way to measure achievement in content areas. Five tables present study data. (Author/SLD)