This paper examines quantile dependence between international stock markets and evaluates its use for improving volatility forecasting. First, we adopt the cross-quantilogram, a correlation statistic of quantile hit processes, and analyze quantile dependence and directional predictability between the US stock market and stock markets in the UK, Germany, France and Japan. Second, we consider a simple quantile-augmented volatility model that accommodates the quantile dependence and directional predictability from the US market to these other markets. The quantile-augmented volatility model provides superior in-sample and out-of-sample volatility forecasts. Finally, we set up a generalized quantile-based approach to improve volatility forecasting for a wide class of asset portfolios.