Air was injected through a well in a thin transparent tank filled with saturated glass beads to study how the size and air saturation of air sparging plumes are affected by particle size and gradation; operational parameters such as injection pressure, well depth, injection pressure pulsing; and well outlet configuration. V-shaped air plumes with an apex between 408 and 608 were obtained for all tests. The air pressure required to initiate sparging agreed closely with the sum of the air entry pressure and the hydrostatic pressure, with higher initiation pressures required in the fine and wellgraded beads. Higher air flow rates and air saturations were obtained in coarser beads at a given pressure, and the variation in flow rate was consistent with estimated air permeabilities. Peak average air saturations were 28-56% for the coarsemedium beads, 10% for the well-graded beads, and 8% for the fine beads. Air saturation and the radius of influence increased modestly (<40%) as the normalized injection pressure exceeded 0.1. Radius of influence increased by approximately a factor of two as the well depth increased, but leveled off once the ratio of radius of influence to well depth reached 0.60-1.05. Pulsing of injection pressure had no effect on the initiation pressure, air flow rate, or air saturation, but increased the size of the air plume and the radius of influence slightly (<15%). Well outlet configuration had only a slight affect the radius of influence (<10%), air saturation (<10%), or air flow rate (<12%). Dye testing showed that water surrounding the air plume circulated during continuous and pulsed sparging. However, pulsed sparging resulted in greater and more defined circulation of water within and adjacent to the air plume, which should reduce mass transfer limitations during sparging.