Abstract The existing spatial keyword cover query only considers the relevance of the text and the position, and ignores the temporal information of the geospatial objects. In this paper, we define a new query, time-aware spatial keyword cover query (TSKCQ), which takes into account the textual relevance, positional relevance, and temporal information of the objects. A new cost function is proposed in TSKCQ, which is used to evaluate the user’s satisfaction in time and space under the premise of satisfying textual constraints proposed by the user. With this function, an object set with the best cost function value would be returned by TSKCQ. We propose a TR-tree for indexing the temporal and spatial information of objects. Based on this, we propose an exact algorithm to tackle TSKCQ, in which effective pruning strategies are used. Finally, experiments demonstrate the efficiency of the proposed algorithm.