Statistical experimental designs were applied to optimize cold-active lipase production by the psychrophilic bacterium Moritella sp. 2-5-10-1. First, a Plackett-Burmen design (PBD) was used to evaluate the significant effects of various fermentation parameters. The results indicated that soybean meal, temperature, and Tween-80 had significant influences on lipase production. The levels of these variables were optimized subsequently using central composite design (CCD). A quadratic regression model of cold-active lipase production was built, and verification experiments confirmed its validity. On subsequent scale-up in a 10-L bioreactor using optimized conditions, cold-active lipase production (30.56 U/mL) was obtained. The results clearly indicated that the model was adequate even on a large scale. To our knowledge, this is the first report of statistical optimization of cold-active lipase production by a psychrophilic bacterium.