As an alternative to trial-and-error methods, a least-squares technique for derivation of finite plate reconstructions has been developed. The method involves minimization sets on corresponding former plate marging segments as a function of the reconstruction (rotation) parameters. Each data value consists of the distance between a point on one former plate margin (as defined by magnetic anomalies and fracture zones, for example) and the chord formed by the nearest two points on the corresponding, rotated former plate margin. The individual distance values are calculated in sets, corresponding to individual anomaly or fracture zone segments, to reduce the likelihood of convergence to a local minimum. This techique is illustrated by derivation of plate reconstructions for anomalies 7 through 13 and 16 in the east central Pacific Ocean, based on published identifications.