The Eight-Parallel Form Anxiety Battery was administered to a class of physiotherapy students on two separate occasions. Differences between persons and test forms were found but there were no differences in scores between occasions, nor was there any interaction between test forms and occasions. Test-retest reliabilities of the forms were found to be satisfactory. Equivalence reliabilities between the forms were found to be greater for measurements made on the same occasion than for measurements made on different occasions; in addition correlations between the forms were higher on the second occasion. In discussing these results it is concluded that more information of a normative nature is required before this promising instrument can be used in non-experimental work.