A technique, which must be used with extreme caution, in which measured values of a potential field (usually gravity or magnetic) at one surface are used to determine the values that field would have at a lower surface. However, the computed field is often erratic and unreliable as a result of the noise contained in the original measurements becoming exaggerated with downward continuation. The method may be useful in resolving anomalies which overlap at the surface where measurements are made, provided the depth to which the field is continued is not below that of the causative bodies themselves. If that depth is exceeded, the computed field may deteriorate completely and become meaningless.