Inhibition of the delayed-rectifier potassium channel current, human ether-a-go-go (hERG), by pharmaceutical agents can lead to acquired long QT syndrome and the generation of potentially lethal arrhythmias and sudden death. There remains an unmet need for higher-throughput assays to screen compounds in preclinical development for the potential to block hERG and cause QT prolongation. We evaluated the rubidium efflux assay for its ability to determine block of the hERG potassium channel. hERG-transfected human embryonic kidney-293 cells were cultured on 96-well assay plates and loaded with rubidium ion by incubating in media in which potassium was replaced by 5.4 mM Rb+. Cells were exposed to test compounds and then depolarized with a K+ channel opening buffer containing 50 mM K+. The supernatant was removed, and cells were lysed using 0.1% Triton X-100. Concentration-response curves were generated for test agents by determining the Rb+ efflux using a flame atomic absorption spectrometer. Multiple trials with cisapride yielded 50% inhibitory concentration values between 308.1 +/- 11 nM to 456.3 +/- 24 nM for inhibition of Rb+ efflux and a Z factor of 0.80 +/- 0.07 (n = 5 plates, 12 wells per plate). The values for inhibition of the hERG channel exhibited a rightward shift in potency as compared to those measured using electrophysiological techniques. In addition, we evaluated 19 blinded compounds at 10 microM in the Rb+ efflux assay, and compared results to those using patch clamp electrophysiology and the dofetilide displacement binding assay. The dofetilide displacement binding assay yielded a good correlation with electrophysiological measurements of hERG block. The rubidium efflux assay lacked sensitivity to consistently identify significant channel blockade. In conclusion, the rubidium efflux assay provides a higher-throughput means to identify potent hERG channel blocking agents, but lacks the sensitivity required to accurately determine the potency of blockade.