Correct functioning of the brain requires the orderly wiring of billions of cells during development. A major mechanism that mediates this is the guidance of axons to their targets by extracellular chemical gradients. Although well studied, the computational rules by which even simple guidance decisions are made are not yet understood at a biophysical level. The task of a growing axon to estimate a direction signal is immense, as unavoidable biological noise corrupts the measurement at all levels of processing. Here, we show how the modulation of growth by a balance of positive and negative feedback can explain the remarkable chemotactic sensitivity observed in vitro. We perform a detailed analysis of the experimental data of ref. [1] which characterizes the neurite growth of ∼2500 rat dorsal root ganglia explants in very shallow gradients of nerve growth factor. Constrained by these data, we construct a model chemical signaling system for growth and guidance in the developing brain. In the model, amplification of the gradient signal occurs via paracrine signalling between cell bodies within the ganglion, while robustness is conferred by the dynamics of receptor trafficking. The model gives a unified and quantitative account of experimentally observed behavior, and yields testable predictions with implications for understanding brain development and repair after injury.[1] Mortimer, D. et al (2009) A Bayesian model predicts the response of axons to molecular gradients. Proc. Natl. Acad. Sci. U.S.A, 106(25), 10296-10301.