How does agricultural land use respond to variation in the price of agricultural water? Answering this question is difficult in California where there is no well-functioning market for water. To overcome this challenge, I use variation in groundwater depth over space and time to proxy for the price of water. This makes sense in a setting where groundwater pumping is unregulated, meaning the effective price of pumped groundwater is the energy cost to pump it. I construct a panel of agricultural fields in Fresno County, California from 2008 to 2016, and estimate a fixed effects model to estimate groundwater depth's effect on transition probabilities between different categories of land cover. I find that groundwater depth reduces the likelihood that parcels will be planted to an annual crop, but increases the likelihood of fallowing land. Groundwater depth seems to have a less profound effect on choosing to plant perennial crops.