A variety of surface chemical measurements, including surface ozone, NO, NO2, NOx, peroxyacetyl nitrate, and nonmethane hydrocarbon, were carried out at a rural coastal site, Nopigia, Crete, and at a rural elevated site, Prases, Crete, within the framework of the Photochemical Activity and Ultraviolet Radiation (PAUR) II experiment, which took place in Greece and Italy in early summer (1--31 May 1999). The total ozone exhibited very large fluctuations (larger than 90 Dobson units, 1 DU = 2.69 ¿ 1016 molecules cm-2) during the campaign with maximum values of around 400 DU during 10 May and minimum values of <300 DU during 20 May, while the total aerosol content exhibited a reverse fluctuation as the W-NW flow on 10 May changed to SW flow on 20 May, transporting air from the Sahara. This special atmospheric situation with such distinct differences in total ozone column and total aerosol content between these 2 days provides the opportunity, on the one hand, to investigate the induced changes on local ozone production rates and, on the other hand, to assess the local photochemistry for different atmospheric conditions at both a high elevated site and a low elevated site. Box model calculations constrained by the measurements show that changes of the total ozone column and hence of J(O1D) can influence net ozone production rates at Nopigia but not at Prases due mainly to the differences in the NOx levels between the two sites. The observed surface ozone concentrations at Nopigia display larger diurnal variations and higher ozone levels on 20 May compared to the 10 May, whereas at Prases, no such differences are observed in agreement with the model results. The 20% change in J(O1D) due to the change in total ozone column can produce ~1.3 ppbv of ozone from 0600 to 1200 UT at Nopigia. However, the concurrent effect of increasing absorbing aerosol (Sahara dust) content on 20 May may mask part of the increase in J(O1D) and offsets part of the increase in local ozone production at Nopigia.