Spectral cytopathology (SCP), which couples Fourier transform infrared micro-spectroscopy (FTIR) to methods of multivariate analysis, is an accurate and robust method for the detection of subtle biochemical changes within individual cells for the purpose of providing early diagnostic information. Screening for Barrett’s esophagus (BE) could allow for early detection of this premalignant condition, thus allowing a timely diagnosis and intervention for esophageal adenocarcinoma. The aim of this study was to assess the diagnostic performance of SCP for identifying and classifying disease in individual exfoliated cells from the esophagus. Infrared data were collected from 12 samples diagnosed with Barrett’s Esophagus, 10 normal samples, 5 dysplastic samples and 2 with adenocarcinoma. Infrared data were pre-processed and analyzed using Principle Component Analysis and Linear Discriminant Analysis (PCA-LDA). Results were then subject to an artificial neural-network (ANN) to measure the diagnostic performance of SCP. The ANN was able to differentiate Barrett’s esophagus from normal squamous with 96.2% specificity, 77% sensitivity and 88% overall accuracy.