This research investigates how parameter optimisation in MIG welding affects material properties and weld quality, assessed through flexural strength and weld bead geometry using experimental methods. Structural testing indicates that weld pool imperfections often lead to metal failure, especially in welded joints. Incorrect parameter selection can cause weldability issues and defects, requiring post-treatment. The main goal is to minimise defects and boost material strength in welded parts by using a quadratic regression model to predict optimal process parameters. JIS G3131 hot-rolled carbon steel, commonly used in automotive parts, is chosen for its economic value and weldability. The experimental setup involves MIG butt weld joints following AWS 1.1 standards, with parameters like welding current, arc voltage, and speed, and responses such as bead height, width, penetration, and flexural strength. X-ray radiography detects weld pool flaws, followed by a three-point bending test to assess durability. Experimental design, including a 3 × 3 orthogonal array of the RSM design and ANOVA, evaluates factor contributions. Predicted flexural strength aligns closely with experimental values. Optimal parameter values are confirmed through destructive and non-destructive testing. This research concludes that optimal parameter usage significantly improves weld quality and strength, making it essential in metal fabrication.