Geographic routing performance is severely affected by imprecise positioning. Inaccurate location estimation leads to severe throughput degradation rendering it energy inefficient. This article presents a study of geographic routing received signal strength (RSS) and time of arrival (ToA) localisation. Both techniques are simulated using the linear least square method and maximum likelihood-based Levenberg Marquardt method. This study investigates routing behaviour in terms of loss rate (LR) and energy spent on unsuccessful routing. The study sheds light on the failure percentage and the consequent power wasted owing to loss of connectivity, lack of forwarding options with progress, traffic congestion and location error. More importantly, it attempts to determine which localisation technique leads to more energy consumption and thus a shorter network life. As expected, it is found that geographic routing throughput differs, depending on the level of accuracy in localisation. It is confirmed that for ToA there is higher location accuracy and a smaller LR than for the RSS technique. Furthermore, although with ToA the network wastes less energy on lost packets, the extra energy consumption figures are higher than for RSS because of the localisation method which is more costly in energy.