How does the way an organism functions affect its subsequent development? Bryozoans are colonial animals that capture suspended food particles from water currents they generate using crowns of ciliated tentacles (lophophores). In many encrusting bryozoans the water passes through and then under the lophophores until it exits the colony at "chimneys" where the lophophores spread apart to form an opening. To determine whether these water currents can induce the formation of new chimneys, I augmented the excurrent flow by injecting seawater into the colony. New chimneys began to develop near the site of seawater injection within as little as one to two days. New chimneys rarely began to develop within this time interval at control sites where I did not inject seawater. This shows that fluid flow controls development in an external fluid transport system lacking pipe-like conduits, as has been found in the vertebrate circulatory system, an internal fluid transport system with pipe-like conduits. These fluid transport systems show feedback between the way they function and their own development. This kind of "function-dependent development" should be differentiated from phenotypic plasticity, since the developing system, not the environment, produces the signals that induce morphological change.