Lockheed Building 157 is one of the United States largest experiments in contemporary daylighting. Built in 1983, the five story structure houses 3,000 employees and uses daylight for ambient illumination throughout its 56,000-m2 office interior. A continuously dimmable fluorescent lighting system supplements interior daylight under the control of open-loop ceiling-mounted photosensors. In 1985 Lawrence Berkeley Laboratory (LBL) began a year-long program to measure lighting system performance in the building. Data from this study indicated that architectural features of the building performed admirably, admitting significant daylight to large areas of the open plan interior. Operational savings, however, were limited by inappropriate performance of the control system in many of the buildings lighting circuits. LBL recently completed a follow-up investigation of the lighting systems in Building 157 addressing the interaction between daylight and the lighting control system with the goal of improving control system performance. We modified a 1,700-m2 test zone by relocating the photosensors, attenuating, the photosensor control signal, changing the response pattern of the photosensors, and implementing a LBL-developed calibration procedure. Following these modifications, we installed four data-acquisition systems and collected detailed data describing illuminance and lighting power demand during two week periods in the summer, equinox, and winter seasons. This paper presents a comparison of lighting system performance before and after the LBL modifications. Analysis of the data indicates our modifications were successful in maintaining the interior illuminance at the target of 350 lux with minimal electric energy consumption.