NADPH diaphorase expression in neurones and glial cells was examined in primary cultures of embryonic cerebellum and cerebral cortex with (i) increasing age in culture and (ii) the exogenous application of glutamate. In neurone-enriched cultures from both regions, NADPH diaphorase histochemistry selectively labelled discrete sub-populations of neurones and glial cells. Double labelling of the cultures showed that 2-4% of the cells with a neuronal phenotype were NADPH diaphorase-positive. Although the total numbers of neurones present in the cultures declined with increased age of cultures, there was no change in the percentage of NADPH diaphorase-positive neurones with time. In contrast, the percentage of NADPH diaphorase-positive glial cells increased from around 10% at 7 days in culture to more than 50% after 3 or more weeks in both cortical and cerebellar cultures. The age-related increase in staining was due to a greater number of cells expressing NADPH diaphorase activity rather than increased activity of existing enzyme. There was a strong correlation between the decline in neuronal cell population and the increase in the number of NADPH diaphorase positive glial cells. To determine whether or not there was a relationship between the loss of neurones and the increased expression of NADPH diaphorase in glia, neurotoxicity experiments were performed using glutamate. In both cortical and cerebellar cultures, glutamate had a significant neurotoxic effect, with a 30-50% loss of neurons 24 h after application. There was no preferential survival of NADPH diaphorase-positive neurones over the rest of the population, suggesting that NADPH diaphorase positive neurones are not selectively spared in these cultures. Glutamate had no effect on the survival of glial cells. However, glutamate cause a significant increase in the NADPH diaphorase staining of the glia. As with the aging cultures, this increase was due to an increased number of cells with enzyme activity rather than increase in the intensity of staining. The increase in NADPH diaphorase staining was not related to the expression of GFAP and was independent of the presence of neurones, since glutamate also increased NADPH diaphorase activity in pure glial cultures. In both neurone-enriched and pure glial cultures, the increase in NADPH diaphorase activity was independent of extracellular calcium and was not attenuated by the NMDA receptor antagonist dizocilpine (MK 801). However, the increase in activity could be blocked by dexamethasone. The precise identity of the enzyme responsible for these effects is unknown, but these data are consistent with the NADPH diaphorase activity we observed being due to an inducible astrocytic form of nitric oxide synthase. The strong correlation between the increased glial expression of NADPH diaphorase and decreased neuronal survival in both aging and glutamate-treated cultures suggests that NADPH diaphorase expression in glial cells may be an important factor governing the survival of neurones in culture.