This was the third in a series of studies investigating various aspects of rumble strips. In this study, to determine the effect of rumble strips on the real-world stopping behavior of drivers, we used a radar gun to collect speed data from over 400 vehicles on the approaches to ten intersections selected from a sample of 274 approaches. We found that, after drivers encountered the first set of in-lane rumble strips, they slowed down earlier on real-world approaches with rumble strips than on real-world approaches without rumble strips?the difference was, on average, 2.0 mph to 5.0 mph (depending on vehicle category and type of approach). In addition, speeding outliers were more likely to slow down earlier on approaches with rumble strips. The effect of the presence of in-lane rumble strips on stopping behavior was greater for approaches where the driver?s view of traffic on the major road is obscured on one or both sides of the road. The study suggests that stop-controlled intersections at which cross-traffic is obscured by manmade structures and/or vegetation on one or both sides of the intersection would be good candidates for implementing in-lane rumble strips. It is worth noting that while in-lane rumble strips are likely to reduce crashes, they cannot eliminate them. Some drivers might still run stop signs and others might misjudge gaps.