Airborne measurements of actinic fluxes in the band of NO2 photodissociation and in the band of O3 photodissociation leading to the production of O(1D) were performed over the island of Lampedusa, in the Mediterranean, in May 1999, in conjunction with ground-based observations of atmospheric structure and surface radiation. Three cases of cloud-free sky with different aerosol characteristics were identified for this study: 18 May, when a large aerosol optical depth (0.51 at 415 nm) was measured, produced by Saharan dust particles that reached 7 km altitude; and 25 and 27 May, when low-to-moderate amounts of continental/marine aerosols (aerosol optical depth of about 0.2 at 415 nm), originating from Europe, were present; the aerosol was confined below 3 km altitude in these 2 days. The measured actinic flux profiles are compared with the results of a radiative transfer model simulation; the model results are used to interpret the observations and study the sensitivity of the actinic flux profiles on the aerosol and ozone distributions. The model was implemented to accurately describe the aerosol properties and vertical profile and was initialized with an extended set of observations to provide a detailed description of the atmosphere. A good agreement between measurements and model results is obtained in all the examined cases. Calculations at fixed solar zenith angle show that during the desert dust event the actinic flux in the band of NO2 photodissociation is reduced up to 24% in the lower and middle troposphere, with respect to the continental/marine aerosol cases. The actinic flux reduction increases with solar zenith angle, and the largest reduction occurs at the aerosol peak. To identify the influence of the aerosol vertical distribution, which in the desert dust case is extremely different from the generally assumed climatological profiles, model calculations were performed with the observed and the climatological distributions. The vertical distribution of aerosol particles does not influence the modeled fluxes at the surface, which basically depend on the aerosol column amount. Its detailed knowledge is conversely important to obtain accurate model results in the lower and middle troposphere: When the climatological instead of the measured profile is used, the actinic fluxes are within ¿3% of the measurements in the days of low-to-moderate optical depths; the modeled fluxes largely overestimate, up to 19%, the measurements in the desert dust case. The sensitivity of the actinic fluxes in the O3 + hν → O2 + O(1D) photodissociation region on the ozone distribution was studied by performing model calculations with the measured and the climatological profile. The ozone profile appears to affect the actinic fluxes in the O3 photodissociation region, particularly at small solar zenith angles and near the surface. Because of an increase of the radiation scattering caused by aerosols in the troposphere, the model fluxes at the surface obtained with the climatological profile are 7.7% higher than those obtained with the measured profile for a zenith angle of about 37¿; differences are smaller (2.5%) at 55¿ and decrease with altitude.