The continuation of potential field data from one irregular surface to another, not always horizontal, is often a necessary component within the data processing and interpretation stream. The most common requirement is to reduce field values (or some related component or derivative) to a horizontal plane, to facilitate further quantitative processing. Methods available to continue data comprise two main approaches. The first (source-based) involves calculating a source distribution that produces a fit to the data and can be used to calculate the field at any other point above. The second (field-based) requires no source determinations and deals with only fields but may involve calculating the field on some intermediate surface. Nine different continuation methods were compared (four source based and five field based) through synthetic tests and on real data from a helicopter-borne survey in Yukon, Canada. The preferred methods of Guspi and Hansen are those that do not involve any theoretical or geometric approximations and involve intermediate calculations on a plane or surface close to the observation surface. The Guspi approach is faster, based on using frequency-domain processing, but the Hansen method uses equivalent sources close enough to and consistently below the observation surface so that no low-pass filtering needs to be used.