Five non-steroidal anti-inflammatory drugs, naproxen, fenoprofen, ketoprofen, diclofenac and piroxicam, were separated and analyzed by electrokinetic supercharging in CZE. Three different setups of the ITP technique were assayed for the separation and preconcentration of these five non-steroidal anti-inflammatory drugs. For the setup that gave the best results, we evaluated the influence of different parameters on separation and preconcentration efficiency such as sample pH, concentration of the leading stacker, BGE composition, electrokinetic injection time, composition and hydrodynamic injection of the solvent plug and of the terminating stacker. In the selected setup, the BGE (10 mM Na2B4O7 + 50 mM NaCl in 10% of MeOH aqueous solution) contained the leading electrolyte while the terminating electrolyte, hydrodynamically injected after the sample (50 mbar x 12 s), was 50 mM of CHES. Prior to sample injection at (700 s at -2 kV) a short plug of MeOH (50 mbar x 3 s) was hydrodynamically injected. The results show that this strategy enhanced detection sensitivity 2000-fold compared with normal hydrodynamic injection, providing detection limits of 0.08 mu g/L for standard samples with good repeatability (values of relative standard deviation, %RSD < 1.03%). Method validation with river water samples and human plasma demonstrated good linearity, with detection limits of 0.9 and 2 mu g/L for river water samples and human plasma samples, respectively (as well as satisfactory precision in terms of repeatability and reproducibility).