The objective of most seismic time-lapse studies is to detect rock-property changes in a subsurface formation caused by fluid withdrawal or injection, often by comparing seismic reflection images of the subsurface before and after the operation. Because rock-property changes can affect the amplitudes of seismic reflection events associated with the boundaries of the formation, amplitude anomalies are the usual target of time-lapse experiments. Sometimes, however, particularly in harder, less porous rocks, a seismic amplitude anomaly can be relatively small and difficult to detect. There is a secondary time-lapse effect, however, which may be detectable even in the absence of a significant reflectivity anomaly: the time delay of reflections from layers beneath a formation whose wave propagation velocity has been altered by the pore-fluid change. We have developed a near-surface correction technique for land data, which we call joint raypath interferometry, to specifically enhance and detect small time delays between corresponding events on two or more comparable time-lapse seismic images. We demonstrate the technique first on a numerical model, and then on an actual time-lapse field survey in which a reflection amplitude anomaly is difficult to detect.