Response surface methodology (RSM) is a mathematical and statistical technique useful for the modeling and optimization of processes in which a response of significance is influenced by a number of process parameters called input variables. The major concern faced by fabrication engineers during activated tungsten inert gas (A-TIG) welding is the selection of the optimum combination of input variables for achieving the desired weld bead geometry. In this research work, the optimization of A-TIG welding process parameters for P92 (9Cr-0.5Mo-1.8W-VNb) steel have been carried out using RSM. The design matrix for conducting experiments was generated using the central composite design of RSM. The four input process variables, such as welding current, torch speed, arc gap, and electrode tip angles, are varied at five levels. A-TIG bead-on-plate welds were made on a 10 mm-thick P92 steel plate as per the combination of input parameters given by the design matrix. Weld bead geometry, such as the depth of penetration, bead width, and width of the heat-affected zone, were measured and recorded as responses. Second-order response surface models were developed for predicting the response for the set of given input process parameters. Moreover, the response optimization was carried out for obtaining the maximum depth of penetration, minimum bead width, and target heat-affected zone width using the desirability approach. The validation experiments were carried out on the determined optimized process parameters, and it was found that there was good agreement between the predicted weld bead dimensions and actual values obtained during the experiments.