Quantitative information on fracture density may be used to delineate areas, structures, or lithologies that have a relatively high density of subsurface fractures. A quantitative estimate of subsurface fracture density can be obtained from core data. The spacing of joints at outcrops is commonly observed to be a direct, linear function of the thickness of the jointed bed. Our method for estimating subsurface fracture density is a statistical estimate based on this empirical relation. The probability of a given number of core-fracture intersections is determined for a range of possible fracture densities, after accounting for core diameter variations, bed thickness, geometric corrections necessitated by bedding dip, and variations in fracture pitch. The true fracture density is estimated by finding the fracture density value that yields a calculated number of fractured beds equal to the observed number of fractured beds. The technique was tested by measuring fracture density in the field at 2 outcrops. Hypothetical cores were constructed on photographs of the outcrops, and estimates of fracture density from the core data yielded values in agreement with those obtained by field measurement. The core data required for estimation of fracture density are core diameter, bed thickness, number of beds (fractured and unfractured), and the angles between core axis, bedding, and fractures.--Modified journal abstract.