One of the key functions of wood is hydraulic conductivity, and the general physical properties controlling this are well characterized in living plants. Modern species capture only a fraction of the known diversity of wood, which is well preserved in a fossil record that extends back over 400 million years to the origin of the vascular plants. Early fossil woods are known to differ in many key respects from woods of modern gymnosperms (e.g., tracheid size, secondary wall thickenings, lignin chemistry, cambium development) and recent discoveries are shedding new light on the earliest stages of wood evolution, raising questions about the performance of these systems and their functions. We provide an overview of the early fossil record focusing on tracheid morphology in the earliest primary and secondary xylem and on cambial development. The fossil record clearly shows that wood evolved in small stature plants prior to the evolution of a distinctive leaf-stem-root organography. The hydraulic properties of fossil woods cannot be measured directly, but with the development of mathematical models it is becoming increasingly feasible to make inferences and quantify performance, enabling comparison with modern woods. Perhaps the most difficult aspect of hydraulic conductance to quantify is the resistance of pits and other highly distinctive and unique secondary wall features in the earliest tracheids. New analytical methods, in particular X-ray synchrotron microtomography (PPC-SRμCT), open up the possibility of creating dynamic, three-dimensional models of permineralized woods facilitating the analysis of hydraulic and biomechanical properties.