A more detailed measurement instrument for dental anxiety has been developed. To investigate the dimensionability of the response, two instruments were administered to 452 persons. The first set of items (33) concerned stimuli which might prove fearsome to patients. Some of these were dental specific, others were not. A second instrument (20 items) elicited the anxiety symptoms experienced. By comparing separate Kaiser-Caffrey Alpha Factor Analyses two factors were identified within each instrument. These were labelled "generalized" phobias and "dental specific" phobias; and "endogenous" and "exogenous" (anticipatory) anxiety symptoms. Four scales were constructed using simple summations of highly loaded variables. The scales are quite reliable. Although correlations among the scales proved to be significant they are all quite low, indicating that the four scales give more or less non-redundant information about the patient. The scale had discriminating validity in differentiating between patients who acknowledge experiencing spontaneous panic/anxiety attacks and those who did not. The data confirm that dental fears are relatively independent of other generalized anticipatory fears. The results cast doubt on previously held assumptions regarding the unidimensionality of anxiety. They are consistent with a new classification of anxiety and phobic disorders presented elsewhere. The implications for research and patient management are discussed.