Background/Aims: The Framingham Risk Score (FRS) and the ACC/AHA Pooled Cohort Score (PCS) are widely used in clinical practice to guide individual patient care decisions. However, these risk scores have been estimated and validated mostly using data from longitudinal cohort studies; their performance when applied to patient data extracted from electronic health records is less well established. Methods: Risk factor data were obtained from the electronic medical record and insurance claims of 84,116 adults receiving care at a large health care delivery and insurance organization from 2001 to 2011. We assessed calibration and discrimination for four risk scores: the published versions FRS and PCS, and versions obtained by refitting the FRS and PCS using Cox regression models. Population subgroups in which the various models gave highly divergent risk predictions were identified using recursive partitioning techniques. Results: The original FRS was well-calibrated (calibration statistic K=7.4), but the original PCS was not (K=39). Discrimination was similar in both models (C-index C=0.740 vs. C=0.747 for original FRS and PCS). The refitted FRS (K=4.6, C=0.754) yielded better calibration and discrimination than the original FRS; the refitted PCS (K=15.1, C=0.746) was better calibrated than the original PCS. Individual risk predictions differed between original and refitted models for some subgroups, but were similar when comparing refitted models. Conclusion: Both the FRS and PCS are appropriate for use in clinical decision support systems that rely on electronic health data, though it may be advisable to refit the models they are based on using available data from the target population to optimize performance.