An enclosure technique was developed to investigate emission rates of semi-volatile and highly reactive biogenic volatile organic compounds (BVOC) from in-tact leaves and branches. A field-deployable sampling and analysis instrument was developed to simultaneously analyze samples from dynamic vegetation enclosures. Analysis is performed on-site using two stage thermo-desorption gas chromatography (GC) with flame ionization detectors and quadropole mass spectrometry. Four capillary chromatography columns are used to accomplish separation and analysis of C5-C15 hydrocarbons. These techniques were used at a mixed northern hardwood forest (Michigan) and a loblolly pine plantation (North Carolina) during three-summer growing seasons (2003, 2004, and 2005). Basal emission rates and diurnal profiles of isoprene, monoterpenes, (MT) and sesquiterpenes (SQT) were determined for six deciduous and seven coniferous tree species from these sites. SQT emissions from Loblolly pine were found to dominate BVOC emissions for hot summertime (>33°C) conditions at the North Carolina site with MT emissions dominating for lower temperatures and nighttime conditions. Isoprene was the dominant BVOC at the Michigan site comprising  95% of the total flux with MT and SQT comprising approximately 4.3% and 0.7% respectively. A canopy model was used to estimate landscape fluxes for the Michigan site for the three compound classes. Typical daytime maximum fluxes (in µgC m-2 hr-1) were 4000-8000 for isoprene, 150-180 for light-dependent MT, 20 for light-independent MT, and <5 for SQT. Modeled isoprene fluxes agreed within 30% of eddy covariance measurements. Cumulative fluxes over a growing season for these compounds were estimated to be 2500, 105 and 7 g m-2 respectively. Fluxes were used to estimate the OH reactivity rates for the Michigan site. Isoprene accounted for over 95% of daytime OH reactivity rates with maximum values of  0.25 s-2. OH reactivity rates at night were two orders of magnitude less than daytime values ( 0.002 s-2) but up to 80% of this nighttime reactivity was due to the inclusion of MT and SQT. Although these compounds have previously been unaccounted for in modeling oxidant chemistry from this site, the fluxes presented here are insufficient to account for recently reported missing OH reactivity or ozone generated OH production.