Ecosystem fluxes of isoprene emissions were measured during the majority of the 2011 growing season at the University of Missouri's Baskett Wildlife Research and Education Area in central Missouri, USA (38.7° N, 92.2° W). This broadleaf deciduous forest is typical of forests common in the Ozarks region of the central United States. The goal of the isoprene flux measurements was to test our understanding of the controls on isoprene emission from the hourly to the seasonal timescale using a state-of-the-art emission model, MEGAN (Model of Emissions of Gases and Aerosols from Nature). Isoprene emission rates from the forest were very high with a maximum of 53.3 mg m−2 h−1 (217 nmol m−2 s−1), which to our knowledge exceeds all other reports of canopy-scale isoprene emission. The fluxes showed a clear dependence on the previous temperature and light regimes, which was successfully captured by the existing algorithms in MEGAN. During a period of drought, MEGAN was unable to reproduce the time-dependent response of isoprene emission to water stress. Overall, the performance of MEGAN was robust and could explain 90% of the observed variance in the measured fluxes, but the response of isoprene emission to drought stress is a major source of uncertainty.