Optimization procedures using a variety of input parameters have gotten a lot of attention, but using three non-edible seed oils of Jatropha (Jatropha curcas), Sesame (Sesamum indicum), and Sweet Almond (Prunusamygdalus dulcis) has a few advantages, including availability and non-food competitiveness. Optimizing a two-stage trans-esterification process using a sodium hydroxide-based catalyst at a fixed catalyst (1.0wt %) and temperature (60 oC) while varying molar ratio (1:3, 1:6, 1:12), time (20–60 min), and mixing speed (500–1000 rpm), to produce optimal responses of yields were studied using response surface methodology (RSM). The optimization solution of molar ratio (1:3), time (40.9 min.), and speed (500 rpm) resulted in an 86.9 % for refined jatropha biodiesel (RJB), the optimization for refined sesame biodiesel (RJB) with molar ratio (1:6), time (41.7 min.), and speed (619 rpm) resulted in an 88.5 %, and the optimization for refined sweet almond biodiesel (RSAB) with the molar ratio (1:3), time (49.359 min.), and speed (500 rpm) resulted in an 88.7 % at the conditions. RJO, RJB, and RSAB had predicted biodiesel yields of 86.9 %, 88.5 %, and 88.7 %, with less than 0.2 % variation, respectively. The characteristics of biodiesel were studied, and the results were determined to meet both ASTM D6751 and EN14214 criteria. The effects of molar ratio, and time on biodiesel yield from their respective oils were important parameters that greatly influenced the yields, but speed only changed the yields marginally. This work has addressed important difficulties influencing mass production of biodiesel such as the utilization of low-cost feedstock such as non-edible vegetable oils, boosting production efficiency through variable optimization of process parameters, and lowering catalyst dosages through catalyst regeneration.