In recent decades, meteorologists have made remarkable progress in predicting the weather, thereby saving lives and considerable sums of money. However, we are way behind when it comes to predicting the effects of environmental change on ecosystems, even when we are ourselves the agent of such change. Given the substantial environmental problems facing our living planet, and the need to tackle these in an ecologically responsible and cost-effective way, we should aspire to develop terrestrial environmental prediction systems that reach the levels of accuracy and precision which characterize weather prediction systems. I argue here that well designed, long-term monitoring programs will be key to developing robust environmental prediction systems.