This paper explores the impact of non-edible biodiesel blends on emission characteristics under various loading conditions in a naturally aspirated four-strokes multi-cylinder diesel engine. A comparative analysis of the emissions characteristics of four non-edible biodiesel (beauty leaf biodiesel, papaya seed biodiesel, stone fruit biodiesel and tomato seed biodiesel) blends (20% vol. = B20) and diesel was performed by varying engine loads (0, 50 and 100%) and speeds (1200, 1800 and 2400 rpm). The aim was to optimise operating parameters such as biodiesel blends, engine loads and speed on engine emissions such as nitrogen oxide (NOx), carbon dioxide (CO2), hydrocarbon (HC), particulate matter (PM), carbon monoxide (CO) and exhaust gas temperature (EGT). A statistical model and analysis of variance (ANOVA) were used to optimise various parameters. At full load condition and 2400 rpm, the minimum and maximum increase in NOx and CO2 emissions were found to be 4.1% to 23.6% and 1.7% to 19% for papaya seed biodiesel PB20 and beauty leaf biodiesel BTL20 respectively. The results reveal that the engine load and speed were the two most imperative parameters that affected emissions (NOx, HC, PM and CO). Both biodiesel blend and the load were responsible for changing the EGT and NOx emissions. While NOx emissions were unaffected by variations in biodiesel blends, load or speed, the CO2 emissions were not affected by the operating parameters. To conclude, papaya seed oil can be a plausible biodiesel feedstock and its diesel blends with varying engine speeds, and loads can provide optimal engine testing characteristics for negotiating NOx, CO2, HC, PM, CO, and EGT concentration.