ABSTRACT: The objective of this experiment was to quantify variation in bovine M. longissimus dorsi tenderness and determine the extent such variation is explained by variation in the ultrastructure of muscles after different postmortem treatments. Eight muscles were hot-boned and aged for 2 d at 2 °C (T1) to achieve very contracted actomyosin crossover and tough beef. Eight carcass sides were tenderstretched for 10 h at 10 °C and a further 38 h at 2 °C (T2) to achieve lengthened actomyosin crossover and tender beef. Both T1 and T2 were compared with conventionally hung carcasses, which underwent similar chilling regimes, C1 (n= 8) and C2 (n= 8), respectively. Measurements of sarcomere length, pH, Warner Bratzler shear force (WBSF), and sensory tenderness were taken, and transmission electron microscopy images analyzed. Variances of attributes were analyzed on Bartlett's test. Variances of the 4 groups were homogenous for all attributes except for pH after 24 h postmortem (with T1 [0.00] having lower variances than C1 = 0.04, T2 = 0.06, and C2 = 0.05) and WBSF after 2 d aging (with T2 [74.33] having lower variances than T1 = 236.76, C1 = 398.82, and C2 = 856.74). The variation in the tenderness of beef was quantified through ultrastructural variation in bovine muscle, with tenderstretched moderately chilled beef having the least variable tenderness as a result of more uniform overlap between actin and myosin filaments. Variation in the eating quality of beef was not reduced by hot-boning with fast chilling or conventional hanging with fast or moderate chilling. The development of the uniformity within filaments of tenderstretched muscle requires further analysis as residual variation remains.