Sweat, as a biofluid that is easy to extract and contains a variety of biomarkers, can provide various types of physiological information for health monitoring. In recent years, research on wearable sensors for sweat sensing has been emerging continuously. Wearable sweat sensing will probably become an alternative method to traditional chemical analysis. This is due to its advantages of portability, non-invasiveness, comfort, and continuous monitoring. Since the inception of this research field, wearable sweat sensors have achieved significant development in terms of materials, structures, systems, and application directions. Research interests are gradually evolving from single biomarker detection to the pursuit of multi-channel, multi-modal system-level architecture. The analysis of physiological signals has also developed from single signal characterization to omics analysis using multiple physiological information sources. Based on the changes mentioned above, this paper mainly introduces the latest researches of wearable sweat sensors from the aspects of strategy, architecture, material, system, data processing, etc., and tries to summarize the trends of sweat sensors. Finally, this paper analyzes the challenges faced by the sensing platform and possible methods for optimization.