A major problem associated with clinical management of cancer is controlling the accompanying pain, and various analgesics are in common use for this purpose. Recent evidence suggests that some of the targets of analgesics, such as ion channels and receptors, may also be involved in the cancer process, thereby raising the possibility that such use of some analgesics may impact upon cancer itself. The main aim of this study was to determine whether gabapentin, a common adjuvant analgesic in current use against cancer-associated neuropathic pain, would affect tumour development and progression in vivo. The Dunning rat model of prostate cancer was used. Strongly metastatic Mat-LyLu cells were implanted subcutaneously into syngeneic Copenhagen rats which were then treated every other day with 4.6-16.8 g/kg gabapentin by gavage. Primary tumourigenesis was monitored daily. Lung metastases were counted and measured after killing the rats 21 days later. Gabapentin had no effect on primary tumourigenesis but produced dose-dependent effects on lung metastasis. Whilst 4.6 g/kg had no effect, 9.1 g/kg gabapentin decreased the number of lung metastases significantly by 64%. In contrast, 16.8 g/kg gabapentin promoted metastasis significantly by 112% and showed a strong tendency to shorten mean survival time. It is concluded that gabapentin prescribed to cancer patients against pain could impact upon the cancer process itself.