Abstract: Considering the operation complexity, analysis and representation difficulty in the multivariable volume data effective rendering, parallel coordinates and interaction are applied to the transfer function design; We introduce the concept of parallel coordinate prime dimension(PPD) and present a method for transfer function design which combines the parallel coordinate with transfer function curves. As the high dimensional transfer function design based on parallel coordinates is not flexible and needs tedious regulation, we extract the feature of PPD using improved intensity-gradient histograms and propose adaptive range selection in parallel coordinate to guide the range adjustment. Mapping the spatial position of sampling points which satisfies the range of each dimension into PPD, extracting the data range and generating transfer function curves in real time. The experimental results show that our approach can reduce the complexity of the design of high dimensional transfer function, improve the flexibility of transfer function regulation, and enhance the rendering result effectively.