Anxiety disorders are common, debilitating, and associated with functional impairment and cognitive biases, such as interpretation biases for negative information. Although efficacious treatments exist for these disorders, many affected individuals do not have access to them. One treatment that has been shown to be efficacious in the last 10 years for anxiety is computerized cognitive bias modification (CBM). Moreover, it is possible to use a specific form of CBM (interpretation bias modification, IBM) to reduce the negative interpretations and increase benign interpretations of ambiguous events, and this in turn can reduce anxiety responses to a stressor. Additionally, IBM interventions can reduce levels of anxiety about anxiety sensations (i.e., anxiety sensitivity, AS). However, studies demonstrating the effect of IBM on AS have not generalized to other response systems (e.g., reduction of physiological responding). To address this issue, in Experiment 1 I tested a single-session IBM intervention to determine if interpretation biases could be modified in 18 participants. In Experiment 2, I used electroencephalogram (EEG) before and after one session of IBM to measure the P300 in 10 undergraduates with high self-reported Anxiety Sensitivity Index (ASI) scores. For both experiments, I hypothesized that after one session of IBM participants would: 1) endorse more benign and fewer negative interpretations of ambiguous scenarios and 2) show a decrease in ASI scores; additionally, for Experiment 2 I hypothesized that IBM participants would: 3) show reduced P300 amplitude for the oddball stimuli. As predicted, there was a significant Time x Valence interaction in the training group for both experiments, such that negative biases significantly decreased and neutral biases did not significantly change. Also, ASI scores significantly decreased in Experiment 1, and there was a significant main effect of Time in Experiment 2 indicating that ASI scores significantly decreased across both conditions. Finally, P300 amplitude for oddball stimuli did not reduce over time. Results from this study suggest that individuals high in AS have interpretation biases that can be modified; however, a multi-session program or more participants may be needed to obtain differential reductions in AS or change of the P300.