The pH of water directly affects growth of mitten crab (Eriocheir sinensis H. Milne-Edwards, 1853) in aquaculture. A prediction model was set up to determine the changing trend of pH value during culture of mitten crabs. The model would help the farmer to take measures in advance to maintain the safety of cultured crabs, when the predicted value of pH is found to cross beyond safe levels. Prediction model of pH is based on the least squares support vector regression (LSSVR) model with chaotic mutation to improve the estimation of the distribution algorithm (CMEDA) to find optimal parameters (γ and σ) of LSSVR. Because these two parameters can significantly affect the performance of the LSSVR, the other three parameter optimisation methods viz., the particle swarm optimisation (PSO) algorithm, the genetic algorithm (GA) and grid search (GS) algorithm were used to compare with the CMEDA algorithm. The calculated mean absolute percentage errors of the results of the four prediction models were 0.4059, 0.6332, 0.9385 and 1.2499%, respectively. The CMEDA-LSSVR model has a higher prediction accuracy and more reliable performance than the other models. The prediction model was used in Xinhua, Jiangsu Province, China and it performed well and helped farmers make decisions and reduce aquaculture risks.