Somatostatin analogues are in clinical use for the diagnosis and treatment of several oncological indications, namely pituitary adenomas and endocrine gastrointestinal tumors. In addition for a variety of malignancies their potential value is being studied. It has been speculated that somatostatin plays a role in the homeostasis of gliomas, and that gliomas could be susceptible to antiproliferative effects of somatostatin analogues. These assumptions were tested in 20 human cell lines derived from malignant gliomas and 4 glioblastoma tissue specimens, which were analyzed for their expression of the five known somatostatin receptor genes (SSTR1-5) and for the receptor function. Using semiquantitative PCR techniques, SSTR2 transcripts were found in all 20 cell lines and 4 glioblastomas, SSTR1 transcripts were detected in 9 cell lines and 4 glioblastomas, and SSTR3 transcripts were noted in 7 cell lines and 1 glioblastoma. SSTR4 and SSTR5 transcripts were only rarely detected. Gene expression profiles in glioblastoma tissue specimens resembled those of the cell lines in quality as well as quantity, with average transcript levels being highest for the SSTR2, followed by SSTR1 and SSTR3. However, when compared to GH3 anterior pituitary tumor cells, the relative amounts of PCR amplified DNA fragments were found to be at least 120 fold lower in glioblastoma cell lines and tumor specimens. Binding studies indicated that glioblastoma derived cells contained only minute amounts of SSTRs. No inhibition of proliferation was observed when 10 selected cell lines were incubated with somatostatin-14 (SST-14) or octreotide (SMS 201-995) at concentrations ranging from 10(-9) M to 10(-6) M, however, the proliferation of two cell lines was weakly stimulated after 6 days of incubation with 10(-6) M octreotide. The activity of adenylate cyclase, stimulated by forskolin, was inhibited by maximally 25% at 10(-6) M SST-14 or octreotide in one of 5 selected glioblastoma cell lines. Somatostatin peptides do not seem to exert anti-proliferative effects on glioblastoma cells and therefore appear to be of no obvious value for glioblastoma therapy. Most likely the amount of cell surface SSTRs is not sufficient to mediate antiproliferative effects. Since it has been described that SSTRs are detectable on most differentiated gliomas as well as astrocytes, it may be speculated that SSTRs may be relevant only in the context of well differentiated cellular programs but lose their significance with progressive dedifferentiation.