Is there any method to study the feasibility of water injection into naturally fractured geothermal reservoirs before implementing liquid injection? How can we identify the difference of water injectability between different reservoirs? Spontaneous water imbibition tests in rocks from The Geysers and in Berea sandstone have been designed to answer these questions. The maximum water saturation and the maximum imbibition rate by spontaneous water imbibition were used to evaluate the feasibility and the ability of water injection into geothermal reservoirs. Reservoirs with high imbibition rates and high maximum water saturation are good candidates for water injection. If there is no imbibition of water into the rocks at all, then the reservoir is not suitable for water injection. The relationships between imbibition rate and time for The Geysers rocks have been measured under countercurrent and cocurrent imbibition conditions, respectively. The measured maximum water saturation by the spontaneous water imbibition in The Geysers rock was about 87.9%, which was greater than that in Berea although the imbibition rate was lower. Experimental results and the continuing practice of water injection in the field showed that The Geysers geothermal reservoir may be appropriate for water injection.