OBJECTIVES Elevated warfarin levels caused by drug-drug interactions (DDIs) increase the risk of gastrointestinal bleeding and vascular accidents. Electronic medical records (EMRs) have been associated with a greater potential for identifying DDIs. We evaluated the influence of EMR use on trends in prescribing potentially harmful drug combinations involving warfarin from a nationwide survey of ambulatory care office visits. METHODS Data were derived from the 2003 – 2006 National Ambulatory Medical Care Survey (NAMCS), a public-use, nationwide probability sample survey of office visits by ambulatory patients to non-Federally employed physicians. Warfarin and interacting drugs prescribed within the same visit were identified. We evaluated the proportion of interaction-related encounters comparing visits with and without the availability of EMRs. Subgroup analyses were performed by combining 2-years of data (2003-2004 and 2005-2006) for more precise estimates. All analyses were weighted to reflect the sampling design of NAMCS to reflect representative ambulatory care use in the U.S. with appropriate variances. RESULTS Total warfarin-related visits increased from 11.8 million in 2003 to 16.6 and 15.4 million in 2005 and 2006, respectively. In 2003-2004 (period 1) the proportion of warfarin visits in which an interacting drug was co-prescribed was 34.4% (95% CI: 21.4% - 47.4%) with EMRs available and 32.0% (95% CI: 27.5% - 36.5%) with no EMRs available. In 2005-2006 (period 2), the proportions were 36.7% (30.9% - 42.4%) with EMRs and 33.5% (28.6% - 38.3%) without EMRs. We observed no significant differences in the proportions of interaction-related encounters between EMR and non-EMR visits in both periods. CONCLUSIONS Between 2003 and 2006, the rates of warfarin prescribing in ambulatory care increased by 27%. The proportions of interaction-related encounters were unaffected by EMR use, due possibly to the limited implementation of EMRs during that period. Further studies as newer data become available are needed.