Hydraulic fracturing plays an important role in developing unconventional natural gas. The large amount of fracturing fluid retention becomes a significant phenomenon in gas fields. Much research has been carried out to explain this mechanism. Imbibition is regarded as one of the important factors and has been investigated extensively. However, the saturation evolution of different types of fluids (liquid, free gas, and trapped gas) has been less researched during imbibition. A porosity experiment combined with an imbibition experiment was conducted to research the fluids-saturation evolution. There are three types of experimental rocks: tight sand, volcanic rock, and shale. The free-gas saturation decreases with the increasing liquid saturation in all samples. However, the sum of these two types of saturation is approximately 100% during imbibition in tight sand. This indicates that the pore space is almost totally filled by liquid and free gas. The sum of these two types of saturation is less than 100% during imbibition in volcanic rock. This indicates that there is trapped gas by liquid. Trapped-gas saturation increases at the early period and decreases at the late period. The sum of these two types of saturation greatly exceeds 100% during imbibition and increases with the imbibition time in shale rocks. This means that there is large amount of extra imbibition liquid. At the same time, the free-gas saturation fluctuates with the increasing liquid saturation. Based on the above results, it can be concluded that tight sand reservoirs have nearly no trapped gas and extra imbibition liquid, volcanic reservoirs have trapped gas and a little extra imbibition liquid, and shale reservoirs have some trapped gas and a large amount of extra imbibition liquid. This research contributes to understanding the fluid saturation evolution during hydraulic fracturing in unconventional natural gas reservoirs.