This study aims to measure the risk attitude of hybrid maize farmers and to analyze its impacts on the decision of crop transformation in the Mekong Delta, using the empirical method of Eckel and Grossman (2002). Risk attitude is measured by the game of choice with real pay-offs. The risk aversion coefficient is determined on the basis of a utility function with a constant partial risk assumption (CPRA). The study uses data collected from a survey of 150 maize farmers in the Mekong Delta. The research results show that the majority of households are riskaverse, accounting for 67.33%. Farmers with neutral attitude toward risk account for 20% and with a risk-taking attitude for 12.67%. In addition, the results of binary logit regression analysis also show that risk-takers are more likely to transform their crop production models. In addition, the transformation decision also depends on the number of laborers in the households, income from hybrid maize production and disease outbreaks.