This study employed miniaturized personal exposure monitors (PEMs) to measure carbon monoxide (CO) in 588 different commercial settings (e.g., retail stores, office buildings, hotels, restaurants) in five California cities. Altogether, 5000 CO observations were made by recording the instantaneous instrument readings at 1-minute intervals as the investigators walked along sidewalks and into buildings. For 11 of 15 survey dates, two investigators walked side-by-side, permitting two adjacent PEMs to be compared. Quality assurance tests for 1706 pairs of values showed a very high degree of agreement. CO levels for indoor commercial settings were similar to those measured outdoors on sidewalks, apparently because the pollutant seeps into the structures from traffic outside. Although indoor levels usually were above 0 ppm, they seldom were above 9 ppm (the National Ambient Air Quality Standard for an 8-hour exposure), unless some indoor source (e.g., enclosed parking garage) could be identified. For example, an office building with high CO levels from its garage was 'hot' in the sense that CO permeated the upper floors, exposing many office workers to concentrations above 9 ppm, well above ambient levels outside. Indoor settings, without their own sources of CO, were sufficiently similar in concentrations to be treated as a class, although levels did vary slightly from date to date. CO levels on outdoor streets did not vary greatly on different sides of the street, on corners and faces of blocks, and intersections.