Photochemistry in young plumes from vegetation fires significantly transforms the initial fire emissions within the first hour after the emissions are injected into the atmosphere. Here we present an investigation of field measurements obtained in a smoke plume from a prescribed savanna fire during the SAFARI 2000 field experiment using a detailed photochemical box-dilution model. The dilution used in the model simulations was constrained by measurements of chemically passive tracers (e.g., CO) near and downwind of the fire. The emissions of the dominant carbonaceous compounds, including oxygenated ones, were taken into account. The field measurements revealed significant production of ozone and acetic acid in the gas phase. The photochemical model simulations also predict ozone production, but significantly less than the measurements. The underestimation of the ozone production in the model simulations is likely caused by shortcomings of our current understanding of ozone photochemistry under the polluted conditions in this young smoke plume. Several potential reasons for this discrepancy are discussed. One possible cause could be the neglect of unmeasured emissions or surface reactions of NO2 with methanol or other hydrocarbons. In contrast to the field measurements, no significant production of acetic acid was simulated by the model. We know of no gas-phase reactions that cause the production of acetic acid on the timescale considered here. Though many processes were well-simulated by the model, there is a need for further research on some key photochemical processes within young plumes from biomass burning and the potential interactions between gas and the particulate phases. These fundamental photochemical processes may also be of importance in other polluted environments.