Proteases are proteolytic enzymes with a wide range of industrial, biotechnological, and medical applications. Due to their importance, proteases have been the subject of many attempts to engineer improved performance, but campaigns to improve activity via directed evolution have been hindered by inefficient analytical techniques and insufficient understanding of sequence-function relationships. Tobacco etch virus protease (TEVp) has mostly been engineered for attributes other than catalytic activity, and most of the past efforts have employed random mutagenesis methods such as error-prone PCR as opposed to targeted mutagenesis. We developed a novel and seemingly generalizable yeast surface display approach that co-displays protease mutants adjacent to substrate on the same Aga2 anchor protein. Enhanced activity mutants are identified by protease cleavage of tethered substrate removing an epitope tag, which empowers flow cytometric isolation of cells with a decrease in anti-epitope antibody signal. This platform was shown to quantitatively differentiate catalytic activity at the single-cell level for TEVp and sortase A. We leveraged this display platform to perform high throughput screens on seven structure-based active site combinatorial libraries created via saturation mutagenesis, and then screened a second-generation library combining the resultant beneficial mutations. Deep sequencing of functional mutants elucidated sequence-function relationships across 34 sites and identified improved multi-mutants. Clonal analysis of a host of recombinant TEVp multi-mutants with purified substrate demonstrated up to 2.9-fold improvement in catalytic efficiency, generally via decreased KM. The novel yeast surface protease/substrate co-display system and the insights gleaned on rational active site library design and the TEVp sequence-function map will aid future protease engineering efforts, and the collection of improved multi-mutants will benefit the biotechnological community in utilizing TEVp in its multitude of applications. One class of application for engineered proteases is physiological release of diagnostic or therapeutic moieties. We introduced a novel extension of synthetic reporters to noninvasively detect abnormal receptor expression. Synthetic reporters have been demonstrated to noninvasively detect a host of diseases via nanoparticles conjugated to reporters via substrate linkers; biomarkers are generated dependent upon a disease-specific enzyme and filtered into the urine. This approach is limited by its reliance on upregulation of disease-specific proteases, but many diseases are characterized by abnormal expression of cell-surface receptors. The new approach harnesses ligand-enzyme fusion proteins to impart exogenous enzymatic activity to tissue with aberrant receptor expression. A mathematical model for epidermal growth factor receptor (EGFR) tumor xenografts in mice demonstrated feasibility of this approach with TEVp-based fusions, suggesting detection down to tumor diameters of 0.28 mm at standard substrate concentrations. Multiple fusions were produced using different enzymes, ligands, and orientations, and binding and catalytic activity was generally well preserved, indicating a modular fusion framework. Demonstrating feasibility with anti-EGFR TEVp-based fusions in an in vitro cellular assay was not consistently successful. However, the following limitations were identified for improvement: high substrate lability, and insufficient fusion-specific product generation due to inadequate catalytic activity – which would motivate protease engineering – or suboptimal fusion linker design that resulted in ineffective projection of receptor-bound fusion’s enzyme component to engage soluble substrate. Together, this work introduced a novel extension of the synthetic reporter concept to quantify receptor expression, and we have demonstrated theoretical in vivo feasibility as well as empirical functionality of the required ligand-enzyme fusions. We have also introduced a novel display platform that can be harnessed for screening combinatorial protease libraries to find mutants with improved catalytic efficiency, which will aid the synthetic reporter approach.