Background: Adoption of electronic health record systems has increased the availability of patient-level electronic health information. Objective: To examine public support for secondary uses of electronic health information under different consent arrangements. Design: National experimental survey to examine perceptions of uses of electronic health information according to patient consent (obtained vs. not obtained), use (research vs. marketing), and framing of the findings (abstract description without results vs. specific results). Setting: Nationally representative survey. Participants: 3064 African American, Hispanic, and non-Hispanic white persons (response rate, 65%). Measurements: Appropriateness of health information use described in vignettes on a scale of 1 (not at all appropriate) to 10 (very appropriate). Results: Mean ratings ranged from a low of 3.81 for a marketing use when consent was not obtained and specific results were presented to a high of 7.06 for a research use when consent was obtained and specific results were presented. Participants rated scenarios in which consent was obtained as more appropriate than when consent was not obtained (difference, 1.01 [95% CI, 0.69 to 1.34]; P < 0.001). Participants rated scenarios in which the use was marketing as less appropriate than when the use was research (difference, -2.03 [CI, -2.27 to -1.78]; P < 0.001). Unconsented research uses were rated as more appropriate than consented marketing uses (5.65 vs. 4.52; difference, 1.13 [CI, 0.87 to 1.39]). Limitations: Participants rated hypothetical scenarios. Results could be vulnerable to nonresponse bias despite the high response rate. Conclusion: Although approaches to health information sharing emphasize consent, public opinion also emphasizes purpose, which suggests a need to focus more attention on the social value of information use.