This study examines variability in the fine particulate concentrations, meteorological variables and traffic counts in five Southern Californian cities during the summer of 2008. The calculated Monin-Obukhov length suggested that near neutral and slightly unstable conditions were present at both street and roof levels. First, mean wind speed, wind direction, and turbulent characteristics, such as the friction velocity, the vertical wind fluctuation, were compared among the five different cites. Second, the nondimensional variability of vertical velocity and temperature as a function of height showed that the data at surface level within the urban canopy had reasonable agreement with the Monin-Obukhov similarity theory. The observed data in general follows the 1/3 law proposed by Roth, however, involved constants were function of surrounding geometry. Both statistical and analytical methods were applied to model roadside fine particulate concentrations and to analyze the impact of background concentrations, meteorological and traffic-related variables. The generalized additive model showed that the background concentration was the most correlated variable with street fine particle concentrations. This indicated that urban mixing dominated the variation of roadside particle concentrations regardless of urban geometry. Traffic-related variables had opposite effects on particle concentrations in Los Angeles (high-rise settlement) and Huntington Beach (two-story commercial strip). The increasing mean wind speed improved the removal of particles in Huntington Beach. In Los Angeles, particle concentrations were inversely proportional to the vertical velocity fluctuation. A simple analytical model suggested by Berkowicz and Hertel was applied for the special case of wind direction perpendicular to the street axis of Los Angeles to model traffic pollution in streets. The estimated contribution of traffic to roadside DTFP concentrations was 3.2 μg/m3, which is approximately 7% of DTFP concentrations. The estimated emission factor from vehicular emissions was 0.07 g/(vehicle mile).