Training on unstable surfaces is thought to increase the activation of the core (trunk) musculature because of the greater demands for stability. While greater activation of the core has been found with equivalent external resistances, the amount of external force produced by a group of muscles has been found to decrease when the same exercise is performed on an unstable surface. Additionally, while training on unstable surfaces is the most common form of instability training, an individual is more likely to encounter an unstable external resistance rather than an unstable surface, outside of the gym setting. It is unclear how the core muscle activation would compare if relative, rather than absolute, resistances were used, or if an unstable load was used rather than an unstable surface. The purpose of this investigation was to determine the effect of both loading mode and surface condition on the amount of weight lifted during a 10 repetition maximum (10-RM) and the corresponding activation of the core musculature, as measured by surface electromyography (sEMG). Twenty recreationally trained adults performed the overhead press under two loading conditions (barbell/stable load and dumbbell/unstable load) while on two different surfaces (exercise bench/stable surface and Swiss ball/unstable surface). For each condition, subjects performed 3 sets of 3 repetitions with a resistance that was equivalent to a previously determined 10-RM for each condition with sEMG electrodes attached to the following muscles: rectus abdominis (RA), external oblique (EO), upper erector spinae (UES), and lower erector spinae (LES). sEMG signals were collected at 1000 Hz, amplified by 1000 mV, and filtered using a band-pass filter between 20-500 Hz. The root mean square of each signal during a 125ms window was then calculated using a computer algorithm. Peak (pRMS) and integrated (iRMS) root mean square values of each EMG signal were averaged across the three trials for each subject. A 2 × 2 factorial ANOVA with repeated measures was used to compared group mean differences between the 4 conditions (α = 0.05). Stable loads resulted in a 15.6% greater 10-RM, and stable surfaces resulted in an 11.4% greater 10-RM. For the RA, there was no effect for load or surface. For the EO there was a main effect for surface, with 12.5% greater pRMS and 16.8% greater iRMS on the stable surface. For the LES and UES, there was a main effect for load, with the stable load requiring 49.4% and 67.1% greater iRMS, respectively. Additionally, for the UES, the stable load resulted in a 50% greater pRMS. There were no significant interactions. The amount of weight lifted declined with increasing instability of either the load or surface. The RA does not appear to play a strong role in stabilizing the trunk during the overhead press. While the other core musculature responded to changes in surface and loading conditions, increased activation appeared to be a function of the amount of weight lifted overhead. The LES did not increase in peak magnitude, suggesting an overall increase in activation rather than at a specific point in time. Greater activation of the core musculature appears to occur by lifting a heavier weight overhead than by lifting a lighter weight overhead either with an unstable load or on an unstable surface.