Trauma to the peripheral nervous system can result in the hyperexcitability of peripheral and central neurons, leading to sensitisation and enhanced responses to noxious stimulation. Alternative splicing of vascular endothelial growth factor A (VEGF) mRNA results in the expression of two isoform families, VEGFxxxa and VEGFxxxb. These splice variants are important modulators of nociception, altering the excitability of primary afferent neurons; VEGFxxxa isoforms are pro- and VEGFxxxb anti-nociceptive. In these studies, we determined the contribution of VEGFR2 to spinal nociceptive processing, at the level of spinal cord. All experiments were carried out in accordance with UK legislation and the Animal (Scientific Procedures) Act 1986. Both VEGF receptor 2 (VEGFR2) and the splice variant isoforms were identified in the rat spinal cord by immunofluorescence. Nociceptive tests of mechanical allodynia (von Frey hair withdrawal threshold) and thermal hyperalgesia (radiant heat withdrawal latency) were performed in male Wistar rats (~300g) or 129OlaHsd mice (~25g) before and one hour after drug administration. Intrathecal injections (10µl, between lumber vertebra 5/6, 29 gauge needle) of VEGFR2 inhibitor PTK787 (mice), VEGF165a (mice), VEGF165b (mice), or anti-VEGF165b (rat) antibody (56/1) were delivered under isoflurane anaesthesia (2% in O2). Drug effects were compared to i.t. saline (PTK787, VEGF165a and VEGF165b) or mouse IgG (56/1) injections. Data were compared by two way ANOVA + Bonferroni tests (*p<0.05, **p<0.01,***P<0.001) and are presented as mean±SEM. Intrathecal injection of the VEGFR2 full agonist VEGF165a (2.5nM) resulted in significant reductions in withdrawal threshold and latency (thresholds: saline 1.98±0.43g vs. VEGF165a 1.09±0.19g,*,**; latencies: saline 8.84±0.65s, VEGF165a 6.04±0.59s; n=4, *), indicating a pro-nociceptive action. In contrast, intrathecal injection of the partial agonist VEGF165b (2.5nM) or the VEGFR2 inhibitor PTK787 (200nM) increased withdrawal thresholds (saline: 1.17±0.05g; PTK787 2.50±0.29g; VEGF165b 2.42±0.31g, ***) and latencies (saline: 5.99±1.01s; VEGF165b 9.14±0.79s; PTK787 16.31±2.19s; n=3, *, **) indicative of an antinociceptive effect. Intrathecal administration of an antibody raised against VEGF165b (56/1) also led to the reduction in withdrawal thresholds (mouse IgG 32.58±0.64g; 56/1 12.67±2.67g ***) and latencies (mouse IgG 11.06±1.41s; 56/1 7.81±0.49s, *, **). Neutralisation of endogenous VEGF was therefore also antinociceptive in the normal animal. These data, together with our findings in primary afferents respectively, show that VEGF isoforms have consistent differential effects on nociceptive processing throughout the pain neuroaxis in vivo; VEGF165a and VEGF165b have pronociceptive and antinociceptive actions respectively.