The computational procedures used to analyze two-way stop-controlled inters
ections were extended in the National Cooperative Highway Research Project
3-46 to account for a number of effects commonly observed at actual unsigna
lized intersections. This paper presents theoretical extensions that can ac
count for commonly observed phenomena, such as two-stage gap acceptance whe
n median storage is available; right-turn "sneakers" at flared minor-street
approaches; non-random arrivals caused by upstream signals; impedance due
to pedestrian crossings; and delay to major-street through vehicles using s
hared left-turn and through lanes. The individual effects are then combined
into an analytical framework suitable for inclusion in the Unsignalized In
tersections procedures of the 1997 "Highway Capacity Manual". (C) 1999 Else
vier Science Ltd. All rights reserved.