There has been limited attempt to use structural equation modelling (SEM) to assess factors influencing public acceptance of modern biotechnology products such as genetically modified (GM) foods. The use of SEM has several advantages over other techniques. It is an advanced multivariate technique that assesses multiple dependence relationships between variables simultaneously, allows the modelling and prediction of relationships between construct variables in a hypothesized manner and has the ability to suggest novel hypotheses originally not considered. It helps to specify hypotheses and operationalize constructs more precisely and ensures the reliability of measures in the testing of hypotheses in ways beyond the averaging of multi-measures of constructs. SEM has the ability to model latent variables, specify errors, correct measurement error, analyze covariance structures and construct complete theories simultaneously. In cases involving complex issues such as public acceptance of GM foods whereby established models are not available, model generation strategy has been employed. The findings of our study have confirmed that public acceptance of GM foods is a complex issue and should be seen as a multi-faceted process. However when using model generation strategy, it must be acknowledged that the resulting model is in part data driven. In this paper the development of model for public acceptance of GM foods using AMOS version 5.0 software with maximum likehood function will be described as well as the problems and limitations that has been encountered and the measures that has been taken to ensure the establishment of a valid model that does not contradict with existing theory. The findings in this paper will be useful to give some insights in the development of a complex yet well fitting structural equation model using the model generation strategy.