Marine science is a discipline developed on the basis of continuous observation data accumulation, and major breakthroughs in marine science development history are inseparable from the updating of ocean datasets. The time extension of existing gridded datasets has become the main form of updating ocean data products. This review summarizes the current development of gridded marine environment datasets. First, the historical development of ocean observations is divided into three stages: an initial accumulation period dominated by sparse observations, a rapid growth period guided by international observation programs, and a high-quality development period driven by data assimilation and ocean reanalysis. Starting from the three key elements of temperature, salinity, and ocean current, we focus on the global gridded ocean environment datasets published and updated internationally in recent decades, including six flow field datasets, such as HYCOM and OFES, and ten thermohaline datasets, such as Argo and IAP. Based on previous studies, the sources, characteristic information, advantages, and disadvantages of these datasets are briefly reviewed to provide a reference for marine scientists. Finally, the future development direction and research focus of ocean gridded datasets are discussed.