Narrow bracketing in combination with loss aversion has been shown to reduce individual risk taking. This is known as myopic loss aversion (MLA) and has been corroborated by many studies. Recent evidence has contested this notion, indicating that MLA's applicability is confined to highly artificial settings. Given the impact of these findings, we reevaluated the evidence on MLA with substantially higher statistical power than in almost all previous studies. We systematically modified the seminal study design by Gneezy and Potters {[}Gneezy U, Potters J (1997) An experiment on risk taking and evaluation periods. Quart. J. Econom. 112(2):631-645] to include five key adjustments. These involved downscaled returns, return compounding, and extended investment horizons. Our results, which are highly robust to analytical heterogeneity, consistently document the presence of MLA across all experimental conditions.