The green fluorescent protein (GFP) of Aquorea victoria is fluorescent when expressed as a recombinant protein in eukaryotic cells and has been used as a convenient marker of gene expression in vivo. It has also been used as a marker of the intracellular targeting of recombinant fusion proteins (part GFP, part protein of interest) which have been transiently expressed in eukaryotic cells grown in tissue culture. Thus, the use of GFP has proved a useful tool to study intracellular events in real-time. However, some transiently transfected cells fail to express, or localise correctly, the GFP-tagged protein. Therefore the production of stable cell lines expressing GFP-tagged integral membrane proteins may be essential for long-term studies. The generation of stably transfected eukaryotic cells expressing an integral membrane protein with a known, but poorly characterised intracellular trafficking pathway, would provide useful reagents for future, more precise, analysis of that pathway. TGN38 is a type I integral membrane protein which cycles between the trans-Golgi network (TGN) and cell surface; at steady state it is localised to the TGN. As such, TGN38 is an ideal candidate for tagging with GFP. We have generated cDNA constructs encoding ratTGN38 tagged at either the N- or C terminus with GFP. Transiently transfected rat (NRK) cells expressed active fluorophore, but failed to show correct localisation of the fusion protein. In contrast, both constructs are appropriately localised in stably transfected NRK cells and both are fluorescent. Furthermore, the recombinant GFP-tagged proteins and the endogenous TGN38 molecules show identical responses to drugs and temperature blocks known to perturb intracellular morphology and membrane traffic pathways. In fact morphological changes to the TGN induced by brefeldin A were observed at earlier time points than had been described previously using immunofluorescence analysis of fixed cells, thus validating the use of in vivo, real-time analysis of GFP-tagged proteins. In addition, we show that (in contrast to the situation in COS cells) elevated expression of ratTGN38 in NRK cells does not lead to a fragmentation of the TGN; this has implications for the role which TGN38 is playing in the maintenance of the morphology of the TGN. The data we present demonstrate that: (i) it is possible to generate stable cell lines expressing integral membrane proteins tagged with GFP; (ii) the GFP tag remains fluorescent when expressed on either the cytosolic or the lumenal side of all membranes of the secretory pathway up to and including that of the TGN; (iii) the GFP tag does not interfere with the transport of TGN38 along the secretory pathway or its retention in the TGN; (iv) GFP remains fluorescent in cells which have been processed for immunofluorescence analysis (using either paraformaldehyde or methanol fixation); and (v) TGN38 plays a role in maintaining the morphology of the TGN. Thus, stably transfected cells expressing GFP-tagged integral membrane proteins can be used as effective tools for the real-time study of intracellular morphology and membrane traffic pathways in eukaryotic cells.