Robotic reconnaissance ("recon") has the potential to significantly improve scientific and technical return from lunar surface exploration. In particular, robotic recon can be used to improve traverse planning, reduce operational risk, and increase crew productivity. To study how robotic recon can benefit human exploration, we recently conducted a field experiment at Black Point Lava Flow (BPLF), Arizona. In our experiment, a simulated ground control team at NASA Ames teleoperated a planetary rover to scout geology traverses at BPLF. The recon data was then used to plan revised traverses. Two-man crews subsequently performed both types of traverses using the NASA "Lunar Electric Rover" (LER) and simulated extra-vehicular activity (EVA) suits. This paper describes the design of our experiment, presents our preliminary results and discusses directions for future research.