During repetitive movement, low-back loading exposures are inherently variable in magnitude. The current study aimed to investigate how variation in successive compression exposures influences cumulative load tolerance in the spine. Forty-eight porcine cervical spine units were randomly assigned to one of six combinations of mean peak compression force (30%, 50%, 70% of the predicted tolerance) and loading variation (consistent peak amplitude, variable peak amplitude). Following preload and passive range-of-motion tests, specimens were positioned in a neutral posture and then cyclically loaded in compression until failure occurred or the maximum 12 h duration was reached. Specimens were dissected to classify macroscopic injury and measurements of cumulative load, cycles, and height loss sustained at failure were calculated. Statistical comparisons were made between loading protocols within each normalized compression group. A significant loading variation x compression interaction was demonstrated for cumulative load (p = 0.026) and cycles to failure (p = 0.021). Cumulative compression was reduced under all normalized compression loads (30% p = 0.016; 50% p = 0.030; 70% p = 0.020) when variable loading was incorporated. The largest reduction was by 33% and occurred in the 30% compression group. The number of sustained cycles was reduced by 31% (p = 0.017), 72% (p = 0.030), and 76% (p = 0.009) under normalized compression loads of 30%, 50%, and 70%, respectively. These findings suggest that variation in compression exposures interact to reduce cumulative compression tolerance of the spine and could elevate low-back injury risk during time-varying repetitive tasks. (C) 2019 Elsevier Ltd. All rights reserved.