This work constructs multiple regression equations of surface ozone concentration based on non-linear combinations of high temporal frequency and multi-year measurements of air pollutant concentrations (PM2.5, CO, NO2, SO2) and remotely sensed ultraviolet Index (UVI) in nine different urban regions in China. These nine regions all have different emissions profiles, economic drivers, and climatology, allowing a more rigorous investigation of the factors most responsible to local surface ozone. The results show a good fit of ozone can be made temporally (including many peaks and troughs), under conditions ranging from relatively clean through polluted, with minimum and maximum bounds on the goodness of the fit usually in the range from 5 to 130 ug/m3. Overall, the results demonstrate significant differences in terms of the most important driving factors in the different cities, with UV radiation being most important in all cities, followed by CO, PM2.5, and NO2 or a combination, depending on each individual city. The performance of the ozone prediction and real measurements under both clean and polluted conditions of PM2.5 or CO mass concentrations are further explored and found to match very well in Xi’an and Beijing. Discussion is presented and supported to quantify insights into why solar ultraviolet radiation coupled with easier to measure longer-lived air pollutants contribute a significant amount to surface ozone is possible, all without needing to necessarily at first order wade into the extremely complex chemistry and physics involved with boundary layer meteorology and VOC chemistry.