Overexposure to welding fume constituents, particularly manganese, is of concern in the construction industry due to the prevalence of welding and the scarcity of engineering controls. The control effectiveness of a commercially available portable local exhaust ventilation (LEV) unit was assessed. It consisted of a portable vacuum and a small bell-shaped hood connected by a flexible 2 inch (50.8 mm) diameter hose, in both experimental and field settings. The experimental testing was done in a semienclosed booth at a pipefitter training facility. Five paired trials of LEV control vs. no control, each approximately 1 hr in duration and conducted during two successive welds of 6 inch (152.4 mm) diameter carbon steel pipe were run in random order. Breathing zone samples were collected outside the welding hood during each trial. In the field scenario, full-shift breathing zone samples were collected from two pipefitters welding carbon steel pipe for a chiller installation on a commercial construction project. Eight days of full-shift sampling were conducted on both workers (n = 16), and the LEV was used by one of the two workers on an alternating basis for 7 of the days. All samples were collected with personal sample pumps calibrated at 2 L/min. Filter cassettes were analyzed for total particulate and manganese concentration by a certified laboratory. In the experimental setting, use of the portable LEV resulted in a 75% reduction in manganese exposure (mean 13 microg/m(3) vs. 51 microg/m(3); p < 0.05) and a 60% reduction in total particulate (mean 0.74 mg/m(3) vs. 1.83 mg/m(3); p < 0.05). In the field setting, LEV use resulted in a 53% reduction in manganese exposure (geometric mean 46 microg/m(3) vs. 97 microg/m(3); p < 0.05) but only a 10% reduction in total particulate (geometric mean 4.5 mg/m(3) vs. 5.0 mg/m(3); p > 0.05). These results demonstrate that LEV use can reduce manganese exposure associated with welding tasks in construction.