Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) phosphorylated at Thr^286/287 (pCaMKII) is critically involved in the induction of long-term potentiation at many glutamatergic synapses. However, pCaMKII may also be involved in the maintenance of synaptic strength. If so, it would be expected that most glutamatergic synapses contain basal levels of pCaMKII. We performed postembedding immunogold labelling of single and serial ultrathin tissue sections using a polyclonal anti-pCaMKII antibody (Promega) to examine postsynaptic levels of pCaMKII at presumed nociceptive and low threshold primary afferent synapses in the dorsal horn of naive, rapidly perfusion-fixed rats (n = 3 rats; the rats were anaesthetised by 50 mg/kg pentobarbitone i.p. 30 min prior to perfusion fixation). In all examined synaptic populations pCaMKII immunolabeling was enriched in the postsynaptic compartment at ~10-60 nm from the postsynaptic plasma membrane. The large majority of synapses formed by peptidergic (containing substance P and calcitonin gene-related peptide, as assessed by immunolabelling of an adjacent section) and non-peptidergic (morphologically identified) presumed nociceptive primary afferent fibre terminals exhibited immunolabelling of their postsynaptic density. Although nearly all synapses formed by presumed, morphologically identified, low-threshold mechanosensitive fibres terminating in laminae IIi-IV were also immunolabelled in at least one of three serial sections, the immunoreactivity (gold particles/μm of postsynaptic membrane) over such synapses was 48% lower than over non-peptidergic nociceptive primary afferent synapses (P = 0.002; Student’s t test, two-tailed). A large variability was observed in immunolabelling between serial sections through individual postsynaptic densities, indicating heterogeneity in pCaMKII contents between subdomains of the postsynaptic density. Suppressing neuronal activity using lidocaine during perfusion fixation did not result in lower levels of pCaMKII immunofluorescence in the dorsal horn, suggesting that the basal pCaMKII immunolabelling observed was not an artifact of aldehyde fixation. Our results suggest that constitutive pCaMKII is available at most synapses, thus potentially contributing to the preservation of synaptic efficacy. Further, the differences in pCaMKII levels between high- and low-threshold primary afferent synaptic populations indicate that autophosphorylation of CaMKII at Thr^286/287 depends not only on the degree of presynaptic activity, but probably also relies on properties of the postsynaptic neuron and/or on heterosynaptic mechanisms.