The value of a pig carcass is largely based on its lean meat percentage (LM%). The ultimate method of accurately measuring the lean meat percentage is a full dissection of the carcass. However, this is both time consuming and expensive. Lean meat percentage prediction equations have been established for various ‘probes’ and indicator cuts. Although these prediction equations have low (less than 2.5) Root Mean Square Errors of Prediction (RMSEP) and high R_a2 values, research continues to investigate other methods to produce equations with greater accuracy and to identify alternative techniques to full dissection. The aim of this experiment was to investigate the accuracy of lean meat percentage prediction equations using data gathered from 1) grading probes, 2) primal cuts and 3) photographs from cross sections of the loin.