Abstract: An isolation method in which an isolation ring is formed to isolate a semiconductor device from other semiconductor devices on a common substrate. The method is suitable for isolating bipolar devices from CMOS or other devices formed on the same substrate and for preventing base current from being injected into the substrate. The method starts with a substrate having a buried sub-collector and a first isolation region that surrounds the portion of the surface to contain the semiconductor device. The first isolation region extends only part of the distance from the surface towards the buried sub-collector. Layers of polysilicon and dual-tone resist are applied, and a first mask is used with an opaque area aligned over the portion of the surface to contain the semiconductor device. The edge of the opaque region terminates above the first isolation region. After exposure, the properties of the dual-tone resist allow a narrow sub-minimum width trench to be removed from the resist to define an isolation ring.