In the present study some experimental parameters for in situ hybridization histochemistry (ISHH) have been analysed using 35S-labelled and alkaline phosphatase-conjugated probes, in order to develop a reproducible double-labelling procedure. We have compared the total exclusion of tissue fixation with tissue sections fixed by immersion in formalin. In addition, the effect of dithiothreitol was assessed both when combining radiolabelled and non-radioactive probes on a single tissue section and when the probes were used separately. Hybridization of unfixed tissue resulted in stronger specific labelling and lower background both for radiolabelled and alkaline phosphatase-conjugated probes. No loss in tissue preservation was seen at the light microscopic level after hybridization of unfixed tissue. High concentrations (200 mM) of dithiothreitol strongly suppressed background when using 35S-labelled probes, whereas in the non-radioactive procedure, alkaline phosphatase labelling could only be achieved with very low dithiothreitol concentrations (less than 1 mM). This incompatibility led to a protocol using unfixed tissue sections and a sequential hybridization procedure, with the radiolabelled probe and high concentrations of dithiothreitol in the first step and the alkaline phosphatase-conjugated probe without dithiothreitol in the second step.