The Uintah Basin in northeastern Utah, a region of intense oil and gas extraction, experienced ozone (O<sub>3</sub>) concentrations above levels harmful to human health for multiple days during the winters of 2009–2010 and 2010–2011. These wintertime O<sub>3</sub> pollution episodes occur during cold, stable periods when the ground is snow-covered, and have been linked to emissions from the oil and gas extraction process. The Uintah Basin Winter Ozone Study (UBWOS) was a field intensive in early 2012, whose goal was to address current uncertainties in the chemical and physical processes that drive wintertime O<sub>3</sub> production in regions of oil and gas development. Although elevated O<sub>3</sub> concentrations were not observed during the winter of 2011–2012, the comprehensive set of observations tests our understanding of O<sub>3</sub> photochemistry in this unusual emissions environment. A box model, constrained to the observations and using the near-explicit Master Chemical Mechanism (MCM) v3.2 chemistry scheme, has been used to investigate the sensitivities of O<sub>3</sub> production during UBWOS 2012. Simulations identify the O<sub>3</sub> production photochemistry to be highly radical limited (with a radical production rate significantly smaller than the NO<sub>x</sub> emission rate). Production of OH from O<sub>3</sub> photolysis (through reaction of O(<sup>1</sup>D) with water vapor) contributed only 170 pptv day<sup>−1</sup>, 8% of the total primary radical source on average (primary radicals being those produced from non-radical precursors). Other radical sources, including the photolysis of formaldehyde (HCHO, 52%), nitrous acid (HONO, 26%), and nitryl chloride (ClNO<sub>2</sub>, 13%) were larger. O<sub>3</sub> production was also found to be highly sensitive to aromatic volatile organic compound (VOC) concentrations, due to radical amplification reactions in the oxidation scheme of these species. Radical production was shown to be small in comparison to the emissions of nitrogen oxides (NO<sub>x</sub>), such that NO<sub>x</sub> acted as the primary radical sink. Consequently, the system was highly VOC sensitive, despite the much larger mixing ratio of total non-methane hydrocarbons (230 ppbv (2080 ppbC), 6 week average) relative to NO<sub>x</sub> (5.6 ppbv average). However, the importance of radical sources which are themselves derived from NO<sub>x</sub> emissions and chemistry, such as ClNO<sub>2</sub> and HONO, make the response of the system to changes in NO<sub>x</sub> emissions uncertain. Model simulations attempting to reproduce conditions expected during snow-covered cold-pool conditions show a significant increase in O<sub>3</sub> production, although calculated concentrations do not achieve the highest seen during the 2010–2011 O<sub>3</sub> pollution events in the Uintah Basin. These box model simulations provide useful insight into the chemistry controlling winter O<sub>3</sub> production in regions of oil and gas extraction.