Tailored biomaterials with tunable functional properties are crucial for a variety of task-specific applications ranging from healthcare to sustainable, novel bio-nanodevices. To generate polymeric materials with predictive functional outcomes, exploiting designs from nature while morphing them toward non-natural systems offers an important strategy. Silks are Nature's building blocks and are produced by arthropods for a variety of uses that are essential for their survival. Due to the genetic control of encoded protein sequence, mechanical properties, biocompatibility, and biodegradability, silk proteins have been selected as prototype models to emulate for the tunable designs of biomaterial systems. The bottom up strategy of material design opens important opportunities to create predictive functional outcomes, following the exquisite polymeric templates inspired by silks. Recombinant DNA technology provides a systematic approach to recapitulate, vary, and evaluate the core structure peptide motifs in silks and then biosynthesize silk-based polymers by design. Post-biosynthesis processing allows for another dimension of material design by controlled or assisted assembly. Multiscale modeling, from the theoretical prospective, provides strategies to explore interactions at different length scales, leading to selective material properties. Synergy among experimental and modeling approaches can provide new and more rapid insights into the most appropriate structure-function relationships to pursue while also furthering our understanding in terms of the range of silk-based systems that can be generated. This approach utilizes nature as a blueprint for initial polymer designs with useful functions (e.g., silk fibers) but also employs modeling-guided experiments to expand the initial polymer designs into new domains of functional materials that do not exist in nature. The overall path to these new functional outcomes is greatly accelerated via the integration of modeling with experiment. In this Account, we summarize recent advances in understanding and functionalization of silk-based protein systems, with a focus on the integration of simulation and experiment for biopolymer design. Spider silk was selected as an exemplary protein to address the fundamental challenges in polymer designs, including specific insights into the role of molecular weight, hydrophobic/hydrophilic partitioning, and shear stress for silk fiber formation. To expand current silk designs toward biointerfaces and stimuli responsive materials, peptide modules from other natural proteins were added to silk designs to introduce new functions, exploiting the modular nature of silk proteins and fibrous proteins in general. The integrated approaches explored suggest that protein folding, silk volume fraction, and protein amino acid sequence changes (e.g., mutations) are critical factors for functional biomaterial designs. In summary, the integrated modeling-experimental approach described in this Account suggests a more rationally directed and more rapid method for the design of polymeric materials. It is expected that this combined use of experimental and computational approaches has a broad applicability not only for silk-based systems, but also for other polymer and composite materials.