Renewable fuels produced from biomass‐derived sugars are receiving increasing attention. Lignocellulose‐degrading enzymes derived from fungi are attractive for saccharification of biomass because they can be produced at higher titers and at significantly less cost than those produced by bacteria or archaea. However, their properties can be suboptimal; for example, they are subject to product inhibition and are sensitive to small changes in pH. Furthermore, increased thermostability would be advantageous for saccharification as increased temperature may reduce the opportunity for microbial contamination. We have developed a mutagenesis platform to improve these properties and applied it to increase the operating temperature and thermostability of the fungal glycosyl hydrolase Cel7A. Secretion of Cel7A at titers of 26 mg/L with limited hyperglycosylation was achieved using a Saccharomyces cerevisiae strain with upregulated protein disulfide isomerase, an engineered α‐factor prepro leader, and deletion of a plasma membrane ATPase. Using biased clique shuffling (BCS) of 11 Cel7A genes, we generated a small library (469) rich in activity (86% of the chimeras were active) and identified 51 chimeras with improved thermostability, many of which contained mutations in the loop networks that extend over the enzyme's active site. This BCS library was far superior as a source of active and stable chimeras compared to an equimolar library prepared from the same 11 genes. Biotechnol. Bioeng. 2012; 109: 2710–2719. © 2012 Wiley Periodicals, Inc. High concentrations of sugar accumulate during lignocellulose hydrolysis for biofuels; therefore, an elevated operating temperature is critical in reducing contamination risk. To achieve a cellulase (Cel7A) that can operate stably at high temperatures, a new approach for DNA shuffling was developed that featured biased template proportions of 11 Cel7A genes. A streamlined high‐throughput screening platform for fungal cellulases was employed to identify several stabilizing mutations.