Abstract: Atmospheric compound pollution events, primarily those involving PM2.5 and O₃, represent one of the significant challenges faced by China in mitigating atmospheric environmental pollution. The accurate forecasting of atmospheric composite pollutants is essential for implementing effective pollution control and prevention measures. There is considerable uncertainty in forecasting compound pollutant concentrations. However, conventional observations are insufficient to meet the demands for the accurate forecasting of these pollutants. Targeted observations focus on the observational needs of forecasts and present a new strategy to enhance numerical forecasting skills. Currently, targeted observations have been successfully applied to theoretical research and practical field trials for high-impact weather and climate events, significantly improving forecasting skills. Compared to targeted observation research on high-impact weather and climate events, studies on targeted observations for air pollution events began relatively late, and this approach has not yet been implemented in field experiments. This paper reviews the research progress regarding the use of targeted observations for high-impact weather and climate event forecasts, evaluates the application of targeted observation strategies in studies of severe air pollution events, and discusses current challenges. Additionally, it highlights key areas for future research and explores the critical role of targeted observations in improving the forecasting accuracy of atmospheric composite pollutants, aiming at providing scientific support for the precise management of atmospheric composite pollution.