Bifidobacterium pseudocatenulatum G4, a wild strain isolated from infant stools that has previously exhibited probiotic characteristics, was used in this study. The aim of this research was to improve the growth potential of this strain in milk-based medium. An initial screening study using a $2^3$ full factorial design was carried out to identify the impact on biomass production of the various components of the medium which were skim milk, yeast extract, and glucose. Statistical analysis suggested that yeast extract had a significant positive effect on viable cell count whereas glucose had a negative effect. Response surface methodology (RSM) was then applied to optimize the use of skim milk and yeast extract. A quadratic model was derived using a $3^2$ face-centered central composite design to represent cell mass as a function of the two variables. The optimized medium composition was found to be 2.8% skim milk and 2.2% yeast extract, w/v. The optimized medium allowed a maximum biomass of 9.129 $log_{10}cfu/mL$, 3.329log units higher than that achieved with 10% skim milk, which is the amount commonly used. The application of RSM resulted in an improvement in the biomass production of this strain in a more cost-effective milk medium, in which skim milk use was reduced by 71.8%.