The traditional Petersons' Experimental techinique for the study of short-term, memory was modified to investige the effects of the intensity of white noise and the differences of sensory modality on short-term retention and heart-rate responses. Subjects served in this experiment were 48 male and female college students. A list of 36 sequences of superspan digits (9 digits each) were visually or auditorily presented. Each trial, the subject was asked to repeat orally a sequence of digits immediately after stimulus presentation. Then, a white noise of 70 db or 90 db would sound for 4, 8, 12, 16, 20 or 24 seconds, in a form of withins counter-balancing. Upon the termination of the white noise, the subject had to repeat aloud the digits given at the begining of the trial. Response measures were scores of items not correctly recalled at each retention interval and the mean heart-rate responses. Data thus obtained were analyzed by analysis of variance, covariance, and polynomial regression. The findings were as follows: (1) The scores of items not correctly recalled, as well as heart-rate changes, were significantly higher for 90 db Group than those for 70 db Group. This showed that white noise of great intensity has some deleterious effect on one's short-term memory and tends to induce autonomic nervous system arousal. (2) There was no significant difference between Visual Group and Auditory Group so for as forgetting scores are concerned. However, evidence showed that sensory modality may influence degree of learning before forgetting occurs. (3) As Peterson & Peterson (1959) had found, amount of retention decreases as the retention interval increases. Since the percentages of retention seemed higher in this study than those in Petersons' study, it was interpreted as an indication that the white noise might not have entirely kept subjects from self-rehearsal during retention intervals.