There is a current and expanding need to measure surface infiltration rate parameters for stormwater infiltration practices used to mitigate the detrimental effects of land development activities on watershed hydrology. We have developed a falling-head soil surface infiltrometer, termed the modified Philip–Dunne (MPD) infiltrometer, that is inexpensive to construct, easy to use, and requires a minimal water volume per test. Because of these characteristics, many MPD devices can be deployed simultaneously to obtain infiltration rate data at multiple locations within a given infiltration practice. Green–Ampt theory was used to derive the expressions needed for analyzing the falling-head data to solve for the field-saturated hydraulic conductivity (Kfs) and Green–Ampt wetting-front suction (ψ). The accuracy of the analysis was determined using numerical experiments in which falling-head data were generated from a computational solution of the axisymmetric form of the three-dimensional Richards’ equation for homogeneous and isotropic porous media with specified input parameters. The falling-head data were then analyzed using a quasi-analytical procedure, and the resulting values of Kfs and ψ were compared with the input values. The accuracy of Kfs and ψ derived from data acquired using the MPD device was then assessed using physical experiments involving three large barrels packed with different types of sand. The Kfs values obtained for the media in the barrels using an MPD infiltrometer were, on average, 82% of the values obtained from whole barrel falling-head tests. The resulting uncertainty in Kfs values from the MPD infiltrometer is considered to be small compared with the orders of magnitude of variability commonly observed for Kfs values in the field.