In 2005, a seminal paper showed that glioblastoma patients aged 18 to 70, whose tumors have a methylated MGMT promoter have a better prognosis than patients with tumors carrying an unmethylated MGMT promoter. As a consequence of this and several confirmatory studies, routine MGMT testing in the clinical setting was promoted. However, only few centers have indeed implemented routine clinical MGMT testing, mostly due the lack of clear clinical consequence and because of considerable technical issues with the testing itself. Recently published results of trials on elderly patients with malignant gliomas have revived the call for routine MGMT testing for clinical decision making. These studies strongly support that MGMT status is a predictive factor for response to temozolomide treatment in elderly patients with malignant astrocytic gliomas and its use for therapy decisions could improve patient management, avoid treatment toxicities and save costs. However, although a number of different protocols for MGMT testing from routinely collected and formalin-fixed and paraffin-embedded tissue have been suggested, there is still no commonly accepted test method with sufficient analytical performance. Protocols established in high-throughput specialized academic or commercial laboratories may not be easily transferable to less specialized laboratories. Thus, before MGMT testing can be used and recommended for clinical decision making, an adequate test method with confirmed high repeatability and reproducibility needs to be identified. To this end, specifically designed investigations including stringently controlled interlaboratory ring trials are needed. Such studies need to take into account the considerable variation in pre-analytical tissue handling (e.g., tissue fixation conditions) between laboratories.