The purpose of this study was to examine the reliability of normalisation methods used in the study of the posterior and posterolateral neck muscles in a group of healthy controls. Six asymptomatic male subjects performed a total of 12 maximum voluntary isometric contractions (MVIC) and 60%-submaximal isometric contractions (60%-MVIC) against the torque arm of an isokinetic dynamometer whilst surface and intramuscular electromyography (EMG) was recorded unilaterally from representative posterior and posterolateral locations. Reliability was calculated using intra-class correlation coefficient (ICC), relative standard error of measurement (%SEM) and relative coefficient of variation (%CV). Maximal torque output was found to be highly reliable in the directions of extension and right lateral bending when the first of three MVIC contractions was excluded. When averaged across contraction direction, high reliability was found for both surface (MVIC: ICC=0.986, %SEM=7.5, %CV=9.2; 60%-MVIC: ICC=0.975, %SEM=10, %CV=13.7) and intramuscular (MVIC: ICC=0.910, %SEM=20, %CV=19.1; 60%-MVIC: ICC=0.952, %SEM=16.5, %CV=13.5) electrodes. Intramuscular electrodes displayed the least reliability in right lateral bending. The use of visual feedback markedly increased the reliability of 60%-MVIC contractions.