Introduction: Racial and ethnic disparities in ICD use for primary prevention among HF patients have dissipated in recent years. While HF patients who received ICDs non-electively during unplanned or shortly after acute hospitalization were shown to have worse outcomes, racial and ethnic disparities in such non-elective ICD placement have not been described. Methods: We identified older (age≥66) HF patients with ICD placements in 20% random sample Medicare data (2008-2018) and defined ICD placement as: 1) non-elective - during unplanned or after recent (<3 month) hospitalization or 2) elective - during planned hospitalization or as ambulatory procedure. We described the temporal trends in elective vs non-elective ICD placement over time and by racial and ethnic subgroups. Adjusted rate ratios (RRs) for non-elective ICD placement were estimated using multivariable modified Poisson regression. Results: Among 97,925 ICD patients (mean age 76, 67% male, 85% White), Black and Hispanic patients had more comorbidities and were more often dual eligible for Medicaid ( Table ). The rate of non-elective placement for all patients has decreased over time (56% in 2008 to 34% in 2018). Black and Hispanic patients had a higher rate of non-elective placement compared to White patients ( Figure ) which persisted after multivariable adjustment [RR = 1.10 (95% CI: 1.07 - 1.13) for Black and RR = 1.09 (95% CI: 1.04 - 1.14) for Hispanic vs. White patients]. Conclusions: Racial and ethnic disparities in non-elective ICD use persist after adjustment. Future work is needed to understand how these disparities may impact outcomes after ICD placement in racial and ethnic minorities.