Impervious cover signifies buildings, roads, sidewalks, and parking lots of low permeability (assumed nearly zero), and it is a consideration in estimating aquifer recharge in urbanized areas. It is commonly assumed that increasing impervious cover decreases direct recharge from precipitation, but observations of most roads, sidewalks, or parking lots reveal abundant fractures that may permit infiltration. We estimate the secondary permeability of pavements using a double-ring infiltrometer. Linear extrapolation determines the infiltration rate as the water depth approaches zero, which is used as a hydraulic conductivity proxy. Data were collected on pavements in Austin, Texas, at fractures or expansion joints intersected along sixteen 30-meter scan lines. By dividing the sum of the discharges for each fracture by the scan-line area, we determined an equivalent-porous-media hydraulic conductivity. For discrete fractures, these ranged over three orders of magnitude from >10−2 to <10−5 cm/s; scan-line hydraulic conductivities ranged from >10−4 to 10−6 cm/s; permeability along the scan lines tended to be dominated by a few highly conductive fractures; and hydraulic conductivity of the entire paved surface was approximately 5.9 × 10−5 cm/s. By multiplying the paved surface hydraulic conductivity by the time the surface was saturated, we estimate that 21 percent of mean annual rainfall (170 mm) is available as potential recharge. Coupled with enhanced subsurface permeability structure from the installation of utilities and reduction of vegetation and, consequently, evapotranspiration, the net effect of impervious cover on recharge is uncertain, but, in many cases, there is likely either little effect or an increase in recharge from precipitation.