The efficient utilisation of limiting resources (space, materials and energy) is an organisational principle that shapes neural circuits. Consequently a neural circuit motif might be favoured for its economic benefits. I verify this suggestion by examining a circuit motif found in the visual systems of both insects and vertebrates, where photoreceptors responsible for vision in bright light synapse onto second order neurons. Mindful of Marr’s three levels, I review work that shows that this motif performs a necessary computation using an efficient algorithm. I then demonstrate that the implementation brings economic benefits; it makes efficient use of space, materials and energy. The mechanisms employed and their organisation within the circuit reduce the resources required to achieve adequate function by one to two orders of magnitude. I conclude that efficiency shapes neural circuits and observe that four features used to increase efficiency are widely applicable. These are the elimination of unwanted input components prior to vesicle release, the use of non-vesicular mechanisms to apply estimates of population activity, unorthodox computations involving extracellular space, and polyadic synapses in which the transmitter released from a single vesicle drives more than one post-synaptic element, be it neuron or glia.