Automated patch clamp (APC) instruments enable efficient evaluation of electrophysiologic effects of drugs on human cardiac currents in heterologous expression systems. Differences in experimental protocols, instruments, and dissimilar site procedures affect the variability of IC50 values characterizing drug block potency. This impacts the utility of APC platforms for assessing a drug's cardiac safety margin. We determined variability of APC data from multiple sites that measured blocking potency of 12 blinded drugs (with different levels of proarrhythmic risk) against four human cardiac currents (hERG [IKr], hCav1.2 [L-Type ICa], peak hNav1.5, [Peak INa], late hNav1.5 [Late INa]) with recommended protocols (to minimize variance) using five APC platforms across 17 sites. IC50 variability (25/75 percentiles) differed for drugs and currents (e.g., 10.4-fold for dofetilide block of hERG current and 4-fold for mexiletine block of hNav1.5 current). Within-platform variance predominated for 4 of 12 hERG blocking drugs and 4 of 6 hNav1.5 blocking drugs. hERG and hNav1.5 block. Bland-Altman plots depicted varying agreement across APC platforms. A follow-up survey suggested multiple sources of experimental variability that could be further minimized by stricter adherence to standard protocols. Adoption of best practices would ensure less variable APC datasets and improved safety margins and proarrhythmic risk assessments.