Abstract The International Affective Picture System (IAPS) is an extremely valuable resource and among the most widely used stimulus databases. Notably lacking though are separate measures for positive and negative emotion. Thus, researchers cannot distinguish images that evoke a neutral affective state from those that evoke both positive and negative emotion. Further, IAPS ratings are not provided for racial or ethnic groups, leaving it unclear how different groups respond. The current study aimed to aid in image selection by providing information beyond what is currently available. In Study 1, we tested the effects of the Self-Assessment Manikin, a 9-point, pictorial scale used to make the original normative IAPS ratings ( N = 115). When the visual scale was used, ratings were lower and had a restricted range, compared to use of a non-visual, Likert-type, 9-point scale. In Study 2 ( N = 1061), using this non-visual scale, we surveyed a racially and ethnically diverse sample. We provide mean ratings for positive emotion, negative emotion, and arousal, as well as an emotional categorization for each image (positive, negative, or no emotion). Ratings are provided by gender and for Hispanic/Latino/a, East Asian, Southeast Asian, and White, non-Hispanic groups as well as for the combined sample. Preliminary analyses show some group differences in ratings (e.g., East Asian > Hispanic for negative and arousal ratings for one cluster of images), suggesting that demographic variables are related to ratings. The ultimate goal of the current project is to facilitate advances in affective science by providing additional information about IAPS to improve stimuli selection.