Down fabrics are the most natural form of thermal insulation due to their notable breathability. However, the primary issue with down jackets is their limited waterproofness. This study focuses on the down water-repellent process to improve the water-repellent qualities. The first stage discusses the single-quality optimization of process parameters using the Taguchi method. The down water-repellent process parameters for the experiment include the detergent, water-repellent agent, bridging agent and glacial acetic acid. The corresponding down fabric quality characteristics are water-repellent time, elasticity and oxygen number. The second stage was multi-quality optimization using Taguchi-based grey relational analysis. A grey relational grade is determined from the grey analysis. The optimized process parameters simultaneously led to a higher water-repellent, higher elasticity and lower oxygen number then verified through a confirmation experiment. According to experimental verification, the optimal process parameter combinations are: detergent concentration is 0.08% of bath ratio, weight ratio of water-repellent agent to down fabric is 5%, weight ratio of bridging agent to down fabric is 3% and glacial acetic acid concentration is 0.03% of the bath ratio. The corresponding optimal quality characteristics were: water repellency time of 120 min, elasticity of 18.4 cm/30 g and oxygen number of 0.8 mg/100 g. These were compared with the International Down and Feather testing laboratory standards used by the industry, which are water-repellent time of 60 min, elasticity of 14.5 cm/30 g and oxygen number of 10 mg/100 g. This study’s optimized water-repellent process’s qualities are much better than the industrial spec requirements for down fabrics.