An experiment was conducted to determine the variations in concentration of an inert air tracer in a forest where the release point is near ground level. The experiment was not extensive but was planned to be relatively efficient and cost-effective and to provide dispersion data for specific cases of isothermal to neutral atmospheric stability and persistent wind conditions. The experimental site was a 12-m tall broadleaf, evergreen forest of mixed species composition. Measurements of air tracer were obtained within 50 m of the release point and at locations within and just above the forest canopy. Each air sample was collected by spatially averaging through a uniform line sink, approximated by a horizontal length of perforated tube 15-m in length. Air was pulled through each tube at a high rate with a high-volume air pump so that the sample residence time in each tube was about one second. Sulfur hexafluoride (SF6) was released near ground level in the 12-meter forest at a constant rate for 10-20 minute periods. Downstream SF6 concentrations were monitored with portable infrared absorption spectrometers (MIRAN SapphIRe) that were set up to continuously analyze a vapor sample every second. A reference sampling location was 20-m downstream and at the same height as the release, 1.2 m. We measured SF6 simultaneously at other downstream locations to be compared against the measurements at the reference location. In order to avoid having the need for many sample locations, the source was moved upstream and additional trials conducted. 3-D sonic anemometers (Campbell Scientific, Model CSAT3) were placed on a 30-m tower at the 3-m and 21-m height. In addition to the 3-D sonic anemometers, four two-dimensional sonic anemometers mounted on tripods at the 3-m height were spread out from the release point and near to the sample locations.. We computed in real time the heat flux (w’T’) and components of the Reynolds-stress tensor (u’w’, v’w’, u’u’, v’v’, and w’w’), The 2-D sonic anemometers were mounted on 3-m tripods, sampled at 1 Hz, and used for supplementary measurements of mean wind speed and direction at speeds that are very low, typically 0.3 to 0.5 m/s. Vegetation structure measurements were made from four, 100-m2 plots located along a centerline from upstream to downstream of the measurements. Twenty-two trials were conducted in all and most were conducted near dawn to utilize near isothermal conditions and persistent winds. Results were compared against the parallelized finite element model, FEM3, which has been shown to reproduce the complex flow and turbulence structure within and above the canopy. FEM3 predicted the vertical and horizontal distributions of tracer concentration within the expected variations of the SF6 observations for the first 50 m from the source.