Frequency dependent cellular micro-impedance estimates obtained from a gold two-electrode configuration using phase sensitive detection have become increasingly used to evaluate cellular barrier model parameters. The results of this study show that cellular barrier function parameter estimates optimized using measurements obtained from this biosensor are highly susceptible to both time dependent and systematic instrumental artifacts. Based on a power spectral analysis of experimentally measured microelectrode voltages, synchronous, 60 Hz, and white Gaussian noise were identified as the most significant time dependent instrumental artifacts. The reduction of these artifacts using digital filtering produced a corresponding reduction in the optimized model parameter fluctuations. Using a series of instrumental circuit models, this study also shows that electrode impedance voltage divider effects and circuit capacitances can produce systematic deviations in cellular barrier function parameter estimates. Although the implementation of an active current source reduced the voltage divider effects, artifacts produced by coaxial cable and other circuit capacitive elements at frequencies exceeding 1 kHz still remained. Reducing time dependent instrumental fluctuations and systematic errors produced a significant reduction in cellular model barrier parameter errors and improved the model fit to experimental data.