The Epidemic-Pandemic Impacts Inventory (EPII) was developed to assess pandemic-related adverse and positive experiences across several key domains, including work/employment, home life, isolation, and quarantine. Several studies have associated EPII-assessed pandemic-related experiences with a wide range of psychosocial factors, most commonly depressive and anxiety symptoms. The present study investigated the degree to which specific types of COVID-19 pandemic-related experiences may be associated with anxiety and depression risk, capitalizing on two large, independent samples with marked differences in sociodemographic characteristics. The present study utilized two adult samples: participants (N = 635) recruited online over a 4-week period in early 2020 (Sample 1) and participants (N = 908) recruited from the student body of a large Northeastern public university (Sample 2). We employed a cross-validated, least absolute shrinkage and selection operator (LASSO) regression approach, as well as a random forest (RF) machine learning algorithm, to investigate classification accuracy of anxiety/depression risk using the pandemic-related experiences from the EPII. The LASSO approach isolated eight items within each sample. Two items from the work/employment and emotional/physical health domains overlapped across samples. The RF approach identified similar items across samples. Both methods yielded acceptable cross-classification accuracy. Applying two analytic approaches on data from two large, sociodemographically unique samples, we identified a subset of sample-specific and nonspecific pandemic-related experiences from the EPII that are most predictive of concurrent depression/anxiety risk. Findings may help to focus on key experiences during future public health disasters that convey greater risk for depression and anxiety symptoms. (PsycInfo Database Record (c) 2023 APA, all rights reserved).