Abstract.A detailed analysis of OH, HO 2 and RO 2 radical sources is presented for the near field photochemical regime inside the Mexico City Metropolitan Area (MCMA). During spring of 2003 (MCMA-2003 field campaign) an extensive set of measurements was collected to quantify timeresolved RO x (sum of OH, HO 2 , RO 2 ) radical production rates from day-and nighttime radical sources. The Master Chemical Mechanism (MCMv3.1) was constrained by measurements of (1) concentration time-profiles of photosensitive radical precursors, i.e., nitrous acid (HONO), formaldehyde (HCHO), ozone (O 3 ), glyoxal (CHOCHO), and other oxygenated volatile organic compounds (OVOCs); (2) respective photolysis-frequencies (J-values); (3) concentration time-profiles of alkanes, alkenes, and aromatic VOCs (103 compound are treated) and oxidants, i.e., OH-and NO 3 radicals, O 3 ; and (4) NO, NO 2 , meteorological and other parameters. The RO x production rate was calculated directly from these observations; the MCM was used to estimate further RO x production from unconstrained sources, and express overall RO x production as OH-equivalents (i.e., taking into account the propagation efficiencies of RO 2 and HO 2 radicals into OH radicals).Daytime radical production is found to be about 10-25 times higher than at night; it does not track the abundance of sunlight. 12-h average daytime contributions of individual sources are: Oxygenated VOC other than HCHO about 33%;Correspondence to: R. Volkamer (rainer.volkamer@colorado.edu) HCHO and O 3 photolysis each about 20%; O 3 /alkene reactions and HONO photolysis each about 12%, other sources <3%. Nitryl chloride photolysis could potentially contribute ∼15% additional radicals, while NO 2 * + water makes -if any -a very small contribution (∼2%). The peak radical production of ∼7.5 10 7 molec cm −3 s −1 is found already at 10:00 a.m., i.e., more than 2.5 h before solar noon. O 3 /alkene reactions are indirectly responsible for ∼33% of these radicals. Our measurements and analysis comprise a database that enables testing of the representation of radical sources and radical chain reactions in photochemical models.Since the photochemical processing of pollutants in the MCMA is radical limited, our analysis identifies the drivers for ozone and SOA formation. We conclude that reductions in VOC emissions provide an efficient opportunity to reduce peak concentrations of these secondary pollutants, because (1) about 70% of radical production is linked to VOC precursors; (2) lowering the VOC/NO x ratio has the further benefit of reducing the radical re-cycling efficiency from radical chain reactions (chemical amplification of radical sources); (3) a positive feedback is identified: lowering the rate of radical production from organic precursors also reduces that from inorganic precursors, like ozone, as pollution export from the MCMA caps the amount of ozone that accumulates at a lower rate inside the MCMA. Continued VOC reductions will in the future result in decreasing peak concentrations of ozone and SOA in...