This paper provides evidence that volatility is rough using option prices rather than historical high-frequency data. The authors analyze at-the-money implied volatilities of S&P500 options with short maturities (one month) and find that the log-volatility process exhibits scaling properties consistent with a fractional Brownian motion with Hurst parameter around 0.32, confirming earlier results from historical data. They also apply the Medvedev-Scaillet correction to extract spot volatility from implied volatilities across maturities, obtaining similar Hurst parameters (0.33–0.34). The estimated Hurst values are slightly larger than those from historical data (typically ~0.1–0.2), which is explained by a smoothing effect due to the remaining time to maturity. Through Monte Carlo simulations and analytical derivations, the authors show that using implied volatilities as proxies for spot volatility introduces an upward bias in Hurst parameter estimation, with the bias increasing with option maturity. The study demonstrates that rough volatility is a robust feature of financial markets, observable even through option-implied measures.