A successive moving average subtraction method is developed and applied to black carbon measured over 5-min intervals at a downtown location near many small emitters and at a suburban residential site within the urban plume but distant from specific emitters. Short-duration pulses assumed to originate from nearby sources are subtracted from the concentrations at each site and are summed to estimate middle-scale (approximately 0.1-1 km) contributions. The difference of the remaining baselines at the urban and suburban monitors is interpreted as the contribution to the downtown monitor from source emissions mixed over a neighborhood scale (1-5 km). The baseline at the suburban site is interpreted as the contribution of the mixture of black carbon sources for the entire city. When applied to a 24-day period from February and March 1997 in Mexico City, the analysis showed that 65% of the 24-hr black carbon was part of the urban mixture, 23% originated in the neighborhood surrounding the monitor, and only 12% was contributed from nearby sources. These analyses indicate that a fixed-site monitor can reasonably represent exposures in its surrounding neighborhood even when many local sources, such as exhaust from diesel buses and trucks, affect the monitor.