The enhanced pain sensitivity, or hyperalgesia, that often follows tissue injury involves strengthening of synaptic transmission at primary afferent synapses in the spinal cord. However, both the molecular mechanisms that underlie such plasticity and the specific synaptic populations involved are unclear. Using neuronal tracing and immunoelectron microscopy, we have shown that the capsaicin model of hyperalgesia is associated with translocation of GluA1- but not GluA2/3-containing AMPA-type glutamate receptors to synapses formed by C-fibers that lack neuropeptides such as substance P. By contrast, little or no change in AMPA receptor subunits occur at synapses from nociceptive primary afferent fibers expressing substance P. Curiously, autophosphorylated CaMKII, which is strongly implicated in synaptic plasticity at central glutamatergic synapses, shows decreased levels in the postsynaptic density of synapses from non-peptidergic C-fibers but increased levels postsynaptic to substance P- containing nociceptors in the same pain model. Thus, potentiation of non-peptidergic C-fiber synapses by recruitment of GluA1-containing, possibly calcium-permeable AMPA receptors may underlie some aspects of capsaicin-induced hyperalgesia. Furthermore, unlike in proposed LTP mechanisms at other glutamatergic synapses, there appears to be a dissociation between CaMKII autophosphorylation and activity-dependent AMPA receptor trafficking at the two major types of nociceptive primary afferent synapse.