Abstract Aim Hyperspectral imaging (HSI) technology combines imaging with spectroscopy and can be used for the classification of malignant and non-malignant cells. Thereby HSI combined with artificial intelligent algorithms can be used to predict tumor cells in in Barrett’s carcinoma specimens. Methods HSI imaging records light between the visual and near-infrared light (500-1000nm). For a first feasibility study, this technique was used to discriminate between squamous epithelium and esophageal adenocarcinoma and 45 specimens from Barrett’s carcinoma patients were recorded. In 22 of the 45 investigated specimens contained also squamous epithelium. The specimens were fixed routinely after resection in paraformaldehyde, were sliced to 3μm, and were stained by haematoxylin and eosin (HE). A non-parametric supervised classification learning algorithm (K-nearest neighbours (k-NN)) was used for discrimination. Results Barrett’s adenocarcinoma cells were recorded by HSI in all 45 investigated cases. Squamous epithelium and Barrett’s adenocarcinoma cells displayed differences in the absorbance between the wave lengths of 500 to 700 nm. For both, the squamous epithelium and the Barrett’s adenocarcinoma cells, the intra group variances of the investigated specimens were quite low. 333,275 and 74,000 spectra could be measured from Barrett’s adenocarcinoma and from squamous epithelium, respectively. Specificity, sensitivity and precision with a k-NN (k=5) classifier were 0.74, 0.92 and 0.94 for the presence of Barrett’s adenocarcinoma cells. Conclusions HE-stained squamous epithelium and Barrett’s adenocarcinoma cells showed specific spectral alterations, when measured by HSI. These characteristics could be used in the future to develop a computer-assisted algorithm to discriminate semi-automated for tumor cells Barrett’s carcinoma specimens, which will help to foster decision-making support in histopathological diagnosis.