Different disciplines deal with mesoscale problems between the scales of element (micro) and system (macro) in different ways. Usually, coarse-graining or statistical approaches, which are averaging approaches, are used. However, averaging approaches tend to lack or ignore the governing principles ruling phenomena at mesoscales between elements and systems. Mesoscience will enable us to explore the principles common to all mesoscales to correlate parameters between micro- and macroscales. In chemical engineering, understanding dynamic multiscale structures, which are critical to reaction and transport processes, remain a challenge because they are all complicated at mesoscales of three levels: material, reactor and system, making it difficult to solve problems at each level and to correlate these levels. This paper reviews three decades of work on mesoscale phenomena in chemical engineering from principles, modeling, and simulation, through to application and generalization. It is revealed that all mesoscale phenomena in these systems follow a common principle of compromise in competition between dominant mechanisms, and can be generally formulated as a multiobjective variational problem. This leads to recognition of the possibility of an interdisciplinary science, mesoscience, to encompass all mesoscale problems existing between elemental particles and the observable universe. The possibility of realtime simulation of chemical processes based on mesoscale modeling is discussed, focusing on the structural and logical similarity between problem, model, software and hardware. Finally, this review is concluded with the prospects of the emerging mesoscience, the development of multiscale computation, and the possible realization of virtual reality. It is believed these advances will mark a new age of chemical engineering if materialized, subject to the resolution of three mesoscales and bridging of the two gaps between levels. (C) 2014 Elsevier B.V. All rights reserved.