The noninvasive tracking of glioblastoma cancer stem cells (CSCs) in vivo constitutes a prerequisite for the development of CSC-specific therapies. Therefore, as a pilot study to identify CSC biomarkers for clinical magnetic resonance spectroscopy, 10 CSC lines were investigated using high-resolution H-1-nuclear magnetic resonance (H-1-NMR) spectroscopy at 600 and 800MHz (14.4 and 18.8 T) under reproducible in vitro conditions. The spectra were analyzed using principal component analysis (PCA), and spectral regions of high variability were evaluated regarding correlations to stem cell-related properties (clonogenic index and CD133 positivity) and cell death. PCA revealed that duplicates of CSC lines clustered together suggesting a characteristic H-1-NMR pattern of each CSC line. PCA enabled discrimination between samples with high and low clonogenicity, that is, clustering according to one of the hallmarks of stemness in samples with high viability. High/moderate correlations to clonogenicity and CD133 were found in spectral regions with high variability. In contrast, the mobile lipid signal at 1.28 ppm correlated to cell death, but not to stemness, as published previously for neural progenitor cells. In conclusion, our exploratory study demonstrates the correlation of specific resonances within H-1-NMR spectra with stem cell properties and advocates the use of the 1.28 ppm resonance as biomarker for cell death also in CSCs.