Despite the nearly ubiquitous expression of telomerase in almost all types of malignant human tumors, studies have shown widely varying positivity in the highest-grade glioma, the glioblastomas (GBMs), ranging from 26% to 100% of tumors analyzed. We have previously shown significant variability in positive versus negative telomerase expression from region to region within the same GBM. In this study, we hypothesized that application of new quantitative methodology would extend our previous observations and identify whether there is heterogeneity in levels of protein expression even within areas positive for telomerase in high-grade gliomas. Finally, we sought to correlate quantitative telomerase expression with patient outcome and therapeutic response. Quantitative analysis was achieved by polymerase chain-based TRAP assay with phosphorimager analysis and compared with clinical information obtained from 19 patients, most with primary, untreated GBMs. Results showed up to 3-fold variability in telomerase levels across multiple regional samples from the same patient, as well as between patients. In 5 of 6 patients with recurrent tumors who had received intervening radiation therapy or chemotherapy, telomerase was downregulated in the second, post-therapy sample. These data provide in vivo corroboration of recent in vitro experiments showing telomerase downregulation after radiation therapy or chemotherapy treatment of cell lines. Our finding of variability in levels of telomerase expression in GBMs parallels the known heterogeneity of these tumors for histologic features and cell growth-related factors. Statistical analysis showed no relationship between TRAP score and either time to clinical progression or time to death.