Studies from 1997-2016 have identified race as a predictive factor in regard to emergency department (ED) length of stay (LOS), extending across age groups and sexs. However, it is difficult to discern a true bias towards race without confounding factors of resource limitations and/or socioeconomic status (SES). Increased ED wait times lead to delayed care and decreased patient satisfaction, and can impact disparate populations disproportionately. This paper aims to explore if the LOS in an urban ED, when stratified by illness severity, changes based on age, race, and SES. Patient records from 1/2018 - 12/2018 at a level 1 urban trauma center ED, with census of 80k patients, was pulled from the EHR. Patients were stratified to illness severity by using disposition status. Subjects were then categorized by race, age, and SES, using the National Reading System of social grading. Race categories used: Asian, Black, Hispanic, Native American, and White. Age was defined as: Child (0-12), Adolescent (13-18), Adult (19-64), and Geriatric (65+). SES used the five following labels: Classes A, B, C1, C2, and D. Patient zip codes were cross referenced with census income data. Class A represented the top 4% of earners while class D denoted the bottom 20%. Primary outcomes were total LOS in the ED and time to provider (TTP). To assess baseline prognostic strength of each variable, a logistic regression was performed prior to stratification of patient data. Following, patient data was organized by disposition status. The average and median LOS was recorded for each variable. When not accounting for patient disposition, a logistic regression with univariate analysis showed that neither race nor socioeconomic status were independent predictors of LOS. Age was shown to be a significant predictor: with adults as the reference, a child had a log-odds ratio of 0.6, p= 0.001, 95% CI [0.57, 0.63], indicating roughly a 40% reduction in LOS. Geriatric patients had a log odds of 1.52, p= 0.001, 95% CI [1.50, 1.55]. Thus across all groups the typical geriatric patient was waiting up to 50% longer regardless of their disposition status. When patients were placed into three separate cohorts: admitted, discharged, and observed, the prognostic effect of our variables became evident. Age was again predictive, with admitted children having a median LOS (min) of 296 (IQR 215-360) v adults 538 (IQR 335-920), yielding a difference of 242 min, p=0.001. TTP was inversely correlated with admitted adults being seen 19.45 min, p= 0.0093, 95% CI [4.80, 34.11] faster than children. LOS results had an average 62.14 min, p= 0.001, 95% CI [56.38, 67.91] difference between admitted Black and White patients. TTP was not significantly reduced for White patients compared to other races. While there was no difference in TTP, SES results revealed a difference between admitted Class A and D patients for total LOS [395 (IQR 246-713) v 581 (IQR 367-995)]. Disparity between SES classes was 186 min, p=0.0103, three-fold more than the disparity that race accounts for. TTP correlated with age but no other variables. LOS correlated with all variables with race showing less of a disparity than SES. Unlike literature highlighting the influence of race, our data suggests that SES may be a stronger predictor, signifying needing to shift the discussion from institutionalized racism in health care, to one that addresses institutional prejudice toward lower socioeconomic groups.