In nerve tissue the histochemical nicotinamide adenine dinucleotide phosphate-diaphorase (NADPH-d) reaction is considered a suitable marker for nitric oxide synthase (NOS) activity. We have previously shown that the NOS-specific inhibitor L-nitroarginine (L-NNA) can block NADPH-d staining in intermediolateral (IML) neurons of the rat spinal cord; such a reaction might serve as a control for the presence of a NOS-related catalytic activity, i.e., L-NNA-dependent NO synthesis in these neurons. However, L-NNA inhibition of neuronal NADPH-d is inconsistent and is therefore disputed by others. This prompted us to reinvestigate the reaction conditions to provide a standardized protocol for inhibition experiments. In IML neurons of formaldehyde-fixed spinal cord tissue, inhibition of NADPH-d reaction was tested by preincubation of frozen sections with the flavin-binder diphenylene iodonium chloride (DPI, 10 mu M-l mM) which blocked the NADPH-d reaction in a concentration-dependent way, suggesting an inverse relationship of inhibitor concentration and final reaction product generated. Preincubation with the NOS-specific inhibitor L-NNA in glycine-NaOH buffer (pH 8.5-9.5) but not L-nitroarginine methyl ester (L-NAME) revealed a concentration-dependent blocking effect on neuronal NADPH-d comparable to the effects seen with DPI, suggesting the existence of a L-NNA sensitive NADPH-d activity. Blocking with L-NNA (100 mu M-10 mM) was prevented by excess L-arginine (10-100 mM), suggesting competitive binding sites. NADPH-d staining was not inhibited by 7-nitro indazole, another NOS inhibitor. Thus, in formaldehyde-fixed nervous tissue both DPI and L-NNA inhibit the NOS-associated catalytic NADPH-d activity, thereby preventing NADPH-dependent conversion of nitroblue tetrazolium to formazan.