The sensitivity of the extrajunctional membrane of the locust retractor unguis (R.U.) muscle to microiontophoretic application of l-glutamate was compared with that of the extensor tibiae (E.T.) muscle. The distribution of extrajunctional D-receptors for l-glutamate on R.U. muscle was more uniform than on E.T. muscle. Chronic denervation (at 30°C) caused an increase in the mean extrajunctional D-sensitivity of both muscles, complete development of which occurred 12–14 days post-denervation for R.U. muscle and 10–12 days post-denervation for E.T. muscle. The maximum mean extrajunctional glutamate D-sensitivity of denervated R.U. fibres was about 20 times greater than the sensitivity of innervated control fibres, while the increase for denervated E.T. fibres was about 6-fold. For both muscles the increase in glutamate sensitivity was only partially accounted for by an increase in the input resistance of the fibres after denervation. The extrajunctional D-receptors on denervated R.U. muscle were uniformly distributed, but on E.T. fibres there was considerable variation in local glutamate sensitivity. Between 20–28 days post-denervation the extrajunctional sensitivity of both muscles declined at a time when muscle atrophy was evident. The reversal potential of extrajunctional glutamate responses of denervated muscle, determined using voltage-clamp, was +3.5 ± 2.6 mV (mean ± S.D., n = 8), similar to the reversal potential of junctional glutamate responses of both innervated and denervated fibres.