Applying a novel approach based on online query volume data, this study provides the first large-scale portrait of revolutionary nostalgia among the Chinese, undertaking an empirical analysis of how the aggregate level of nostalgia is shaped. For each Chinese province, we use the normalized frequency of searches for red songs on Baidu, the most widely used online search engine in China, to quantify the local level of nostalgia. We find that the evolving trends of nostalgia among the provinces are similar but stratified. The results from the dynamic panel data analysis using the Generalized Method of Moments indicate that revolutionary nostalgia is significantly affected by a set of socio-economic determinants, including GDP per capita, income inequality, social development, legal development and the degree of globalization.