Using a coplanar waveguide with a series gap in conjunction with dielectrophoresis trapping, consecutive S-parameter measurements between 0.5 and 20 GHz were quickly performed with and without a Jurkat cell trapped to compensate for a relatively noisy and drifting background. Based on sixteen measurements repeated on eight live cells and eight dead cells, differences in both return and insertion losses show two distinct distributions indicating either return loss or insertion loss alone can be used to distinguish a live cell from a dead one. Further, since the frequency dependence is generally linear or absent, discrete-frequency measurement (as opposed to sweep-frequency measurement) of return or insertion loss may suffice. If proven statistically by a much larger number of cells, this should greatly speed up the measurement to facilitate its eventual field use.