In this study, the simultaneous determination of caffeine, theobromine and theophylline in tea samples by chemometric approaches using UV spectrophotometry has been reported as a simple alternative to using separate models for each component. Spectra of caffeine, theobromine and theophylline were recorded at several concentrations within their ranges and were used to compute the calibration mixture between wavelengths 200 and 305 nm at an interval of 1 nm was used for data acquisition. Partial least squares regression and artificial neural network were used for chemometric analysis of data and the parameters of the chemometric procedures were optimized. The analytical performances of these chemometric methods were characterized by predicted residual error sum of squares (PRESS), standard error of prediction (SEP) and recoveries (%) and were compared with each other. A series of synthetic solution containing different concentrations of caffeine, theobromine and theophylline were used to check the prediction ability of the partial least squares and artificial neural network. These two methods were successfully applied to real samples, with no interference from excipients as indicated by recovery study results. The results obtained in this investigation strongly encourage us to apply these techniques for a routine analysis and quality control of the teas.