Following the global financial crisis, examining the relationship between economic policy uncertainty and financial market volatility has gained significant importance. This study employs the Markov-Switching Granger Causality approach and quarterly data from 1991-2022 to investigate the nonlinear relationship between monetary policy uncertainty and stock market volatility in Iran. The findings reveal a dual-regime pattern: in Regime 0 (periods of high uncertainty and severe volatility - such as the structural adjustment policies of the 1990s, sanctions in 2013, the US withdrawal from JCPOA in 2018, and the stock market crash in 2020), a bidirectional self-reinforcing causal relationship exists, creating a vicious cycle that intensifies instability in both markets. Conversely, in Regime 1 (periods of low uncertainty and mild fluctuations), the relationship is unidirectional, running solely from monetary policy uncertainty to stock market volatility. Although the intensity of this effect, with a coefficient of 0.98, is considerably lower compared to the high-volatility regime (4.31), it remains statistically significant. These results emphasize the necessity of active management of monetary policy uncertainty, particularly during major shocks, to break this vicious cycle and stabilize the capital market.