"The purpose of this study was to investigate the oxygen cost of running barefoot vs. running shod on the treadmill as well as overground. 10 healthy recreational runners, 5 male and 5 female, whose mean age was 23.8±3.39 volunteered to participate in the study. Subjects participated in 4 experimental conditions: 1) barefoot on treadmill, 2) shod on treadmill, 3) barefoot overground, and 4) shod overground. For each condition, subjects ran for 6 min at 70% vVO (2)max pace while VO(2), heart rate (HR), and rating of perceived exertion (RPE) were assessed. A 2 × 2 (shoe condition x surface) repeated measures ANOVA revealed that running with shoes showed significantly higher VO(2) values on both the treadmill and the overground track (<0.05). HR and RPE were significantly higher in the shod condition as well (<0.02 and <0.01, respectively). For the overground and treadmill conditions, recorded VO(2) while running shod was 5.7% and 2.0% higher than running barefoot. It was concluded that at 70% of vVO (2)max pace, barefoot running is more economical than running shod, both overground and on a treadmill."
Emphasis mine. Why they also didn't test running flat-out is a mystery...
Via Amby Burfoot on Twitter.
P.S. Consider this a follow-up to this post: "Getting Faster Overnight..."