Three different types of kitchen exhaust systems, namely, wall-mounted overhead, island, and downdraft, are closely investigated. Fans are tested for energy consumption and airflow ratings following the procedure given in ASHRAE Standard 51-1985. The velocity profile in the exhaust duct is measured using a pitot tube traverse. There is a slight difference between the measured velocity profile and the profile for fully developed turbulent flow. The airflow rate is obtained by using an average velocity in the duct equal to 87% of the centerline velocity. Characteristic fan curves are plotted for the different exhaust systems. The flow field between the top surface of the range and the bottom of the hood is quafitatively studied by means of a noncontaminating flow visualization system. The effects of adding side baffles, a perimeter slot, and a visor or varying the hood clearance height are studied to show the enhancement of the capturefficiency of the ventilation system using the flow visualization technique. Adding side baffles greatly increases the capture efficiency for overhead hoods; however, little effect is noticed after applying the perimeter slot. The capture efficiency improves by decreasing the clearance height. Downdraft systems, with the exhaust located either at the sides or mid-plane of the range, are effective only for low cooking pots. Their capturefficiency is dramatically reduced by increasing the pot height. The relationship between capture efficiency and airflow rate is important, since it relates directly to the amount of outdoor makeup air required and the energy required to heat or cool it. The airflow velocities are measured quantitatively using an omnidirectional hot wire probe mounted on a manual traversing device. The air velocity magnitude is found to decrease rapidly with increasing distance from the exhaust opening. A simple potential flow model is used to calculate the corresponding air velocities. The theoretical velocity contours agree closely with the measured values. The three ventilation systems are rated for sound by measuring the output noise generated by the farts relative to the ambient sound leveL An increase of 0 to 24 db is reported on a C weighing scale.