Abstract This study evaluated the effects of sodium hydroxide concentration, cooking temperature, and cooking time on the soda pulping of Musa paradisiaca pseudostem using a three-factor Box–Behnken design and quadratic response-surface models. The reported experimental matrix was first checked for consistency with the stated design; the sodium hydroxide center level was corrected from 10% to 14%, and two apparent transcription errors in the design matrix were corrected to 140 °C for Run 6 and 180 min for Run 7 so that the matrix conforms to the stated 3-level Box–Behnken structure. Reanalysis of the supplied response data gave quadratic-model R² values of 0.9876 for pulp yield and 0.9923 for kappa number, with adjusted R² values of 0.9717 and 0.9824, respectively. Sodium hydroxide concentration, temperature, and time were significant linear terms for both responses (p < 0.001), while several quadratic terms were significant. However, lack-of-fit tests were significant for both models (p = 0.0015 for pulp yield and p = 0.0003 for kappa number), indicating that the fitted quadratic equations should not be treated as validated predictive models. A model-based compromise point within the studied region was approximately 13.76% NaOH, 139.04 °C, and 124.60 min, with predicted yield of 54.11% and kappa number of 17.35; this point requires experimental confirmation before being described as an optimum. The results nevertheless demonstrate the strong influence of pulping severity on delignification and pulp retention and support further controlled optimization of banana pseudostem as a non-wood fiber resource. Keywords:Musa paradisiaca, soda pulping, response surface methodology, Box–Behnken design, banana pseudostem, pulp yield, kappa number