The development of powerful muscle function is fundamental to the strength and conditioning process. Optimal load resistance training uses the load that maximises power output to more efficiently achieve this. However, research has shown that factors including measurement method and training status can significantly influence the optimal load. The five experimental studies of this thesis investigated these factors. First, the way in which the positive lifting phase is identified was examined to establish the underpinnings of ballistic resistance exercise preference over traditional alternatives. The results of this study showed that the positive lifting phase of ballistic resistance exercise did not consider the deceleration phase and when this was applied to traditional resistance exercise a greater portion of the positive lifting phase was spent accelerating the barbell. This finding suggested that the assumption of ballistic resistance exercise superiority is theoretically unfounded whilst potentially posing a greater risk of injury. The next three studies established the reliability and suitability of different methods used to measure resistance exercise power output. The second study revealed that the most practically applicable, theoretically sound and reliable method of obtaining power output used the barbell kinematics approach where the acceleration of the barbell was considered but body mass excluded. This may have important implications for field-based methods that are underpinned by this approach. The results of the third and fourth study reinforced the findings of study two. The third study considered whether neglecting horizontal barbell power caused the barbell kinematics approach to underestimate resistance exercise power output, and established that the horizontal contribution did not exceed 2%. The effect of bilateral asymmetries on barbell power output was examined in the fourth study and demonstrated that although ground kinetic side differences reached 21% they were not transmitted to left and right barbell end power outputs, with left and right bar end differences remaining below 4%. The barbell kinematics approach was then used in the fifth study, to show that stronger, more experienced individuals generated greater mean (17 to 35%) and peak (20 to 45%) power outputs and maximised mean and peak power with loads that were considerably less (3 to 15% of IRM less) than their weaker, less experienced counterparts. Training status did not significantly affect power and optimal load reliability. To summarise, measurement methods should not be used interchangeably. The barbell kinematics approach is recommended to obtain resistance exercise power output but the optimal load should be prescribed on an individual athlete basis and routinely monitored for maximum accuracy.