There are billions of digital images, most of which are not accessible to blind screen reader users because they lack quality image descriptions. Increasingly, digital images include representations of people with diverse identities––including people of color, people who are trans or gender nonconforming, and/or people with disabilities. This adds a layer of complexity for image description authors, who may wonder how to find the words to describe these identities with sensitivity, or if they should be mentioned at all. In this webinar, we will briefly review the basics of writing good image descriptions. Then, we will present examples and some starting guidelines for describing a wide range of marginalized identities. This will be a hands-on webinar on a big topic, so come prepared to practice writing your own image descriptions and to have an interactive discussion on the complexities of describing marginalized identities in alt text.