Structure-function relationships are paramount to predicting in vivo muscle function. Muscle architecture is commonly used to predict both active and passive performance. Although these relationships are assumed to be true for active tension, architectural-based predictions of passive tension are unable to capture muscle-specific differences. Passive tension is thought to arise from the resistive stretch of titin and collagen, but the relative importance of these load-bearing proteins and muscle architecture is unclear. The purpose of this work was to investigate the ability of biochemical and architectural parameters to predict active and passive properties of skeletal muscle at different size scales and strains. To robustly evaluate these predictors, functional changes were induced via tenotomy and the morphologic parameters were reevaluated to refine the active and passive models. An approach combining mechanical testing, biochemical assays, and computation modeling was implemented to show that both active and passive performance characteristics can be accurately predicted using a combination of muscle architecture and protein composition under both healthy and pathologic conditions