The present study is an attempt to investigate the effects of testing situation on heartrate responses for students who differ in test anxiety as measured by Test Anxiety Qestionnaire. Twenty-seven high (HA) and 27 low (LA) test anxiety subjects wereselected from 124 male and female university students, to take a simulative "intelligencetest" administered in laboratory. The EKG of each yoked pair (one HA subject and one LA subject) was recorded in six testing sessions which were consisted of varying stress instructions and test items. The results of beat-by-beat analysis of EKG data showed: (1) that there is no significant difference in heart rate responses between HA and LA groups, (2) that the correlation coefficient between test anxiety scores and bpm changes in heart rate is not significantly different from zero, (3) and that significant cardiac acceleration occured for almost all subjects when testing was announced, during the period of test administration, and when "IQ score" was to be reported.These findings support the arguments that the personality measures are not related to physiological indices of autonomic nervous system arousal, and that the heart rate acceleration may voluntarily occur to the "low anxiety" subjects even though the test anxiety is not subjectively perceived or experienced.