Communication among providers is a crucial aspect of patient care. Poor communication is well documented and can lead to erosion of interspecialty relationships and suboptimal patient outcomes.1,2 The Accreditation Council for Graduate Medical Education identifies interpersonal and communication skills as core competencies, setting broad expectations to teach and model ideal communication.3 However, there is currently little evidence to guide teaching about optimal modes of communication. We conducted a mixed-methods study to measure the impact of an interprofessional education program on interspecialty communication at an academic medical center.4 This report describes providers' preferences and practices related to interspecialty communication. Our intervention was a 60-minute interprofessional education program addressing multidisciplinary approaches to diagnosing prosthetic joint infections. Thirty infectious disease (ID) and orthopedic surgery (OS) providers participated (47% [n = 14] trainees, 40% [n = 12] attendings, 13% [n = 4] midlevel providers). A preintervention survey elicited ratings of the perceived effectiveness and use of various modes of communication, including telephone, in-person, consultation notes, e-mail, and pager. Three months after the intervention, we held a focus group of 12 total providers (25% trainees, 58% attendings, 17% midlevel providers) to explore perceptions of communication. In the survey, most providers rated verbal communication (phone or in-person) as “most effective” (87% OS providers, 85% ID providers) and communication via consultation notes as “least effective” (50% OS providers, 69% ID providers). In practice, however, a minority of both groups (35% OS providers, 23% ID providers) reported using in-person communication “often.” In contrast, ID providers used communication via patient chart “often,” unlike OS providers (61% ID providers, 30% OS providers). Although few identified e-mail as effective (7% OS providers, 15% ID providers), this mode was used often (77% ID providers, 53% OS providers). Focus group participants described scheduling constraints as the primary obstacle to verbal communication. They also described communication via patient chart as suboptimal because it was unlikely to be read, for example: “You just have to make the assumption that [OS providers] are not reading…your note,” said one ID physician, and “If the patient is really having troubles [the OS provider] will put the [ID provider's] note up…and look at the last four lines,” said an OS physician. Participants discussed benefits of e-mail communication, including speed, ease of use, and ability to communicate with multiple providers, for example: “It's a lot easier…to shoot back an e-mail than to call or page me and wait,” said an ID physician, and “I [copy] it to my resident…and it won't fall through the cracks,” said an OS physician. However, integrating e-mail communication into the medical record remained controversial. Opinions ranged from opposition to including e-mail communications because “they're not medical records,” according to an OS physician to feelings of “personal discomfort about…[e-mail] communication…outside of the medical record,” said an ID physician. Our study investigated perceptions and practice of communication between 2 subspecialty groups. Although verbal communication was perceived as superior, many providers used notes in patient charts to communicate. Moreover, despite concerns about e-mail communication, it was used frequently. These findings highlight provider ambivalence regarding the most appropriate modes of interdisciplinary communication and suggest a need for further investigation to identify optimal communication practices and ways to integrate these into competency-based education, thereby potentially enhancing interspecialty relationships and patient care.