A method is disclosed for efficiently determining and writing repeatable runout (RRO) compensation value sets for data tracks on a magnetic disk in a disk drive. The disk drive has a magnetoresistive (MR) head having a read element and a separate write element, and a secondary actuator coupled to the end of a primary actuator for adjusting a head skew angle. In the method, the head skew angle is set such that the read and write elements are substantially aligned along the first track. A position error signal for each servo sector is determined over a predetermined number of disk revolutions during track following along a first data track. An RRO compensation value set is calculated for the first data track based on the position error signals. The RRO compensation value set is written for the first data track. A seek operation is then performed to a second data track.