Over the last three decades, the threat to amphibian animals has become more and more serious. Especially, the loss of natural habitats caused by changes to the way land is used by human beings. This issue has hit amphibians extraordinarily hard. Moreover, it is known that amphibians are particularly vulnerable to human activities because they rely on both terrestrial and aquatic habitats for survival. Due to increased development in many areas, concrete structures are often built along water bodies in order to increase the safety of structures. The development of solid banks along streams related with human improvement has become a major issue in Taiwan. Most ecosystems used by amphibians are lakes, ponds, rivers and stream banks. However, there is no related design solution to accommodate the needs of those amphibians. The need to develop the relevant design specification considering protecting the amphibian is very urgent. The objectives of this study were to investigate (1) the relationships between the physical characteristics and climbing ability of toads, (2) whether statistics analysis can be can be utilized to improve the understandings of the actual attributes of amphibian, (3) whether Artificial Neural Network can be used to simulate the climbing ability of toad. The experimental and numerical results showed that there is a strong positive correlation between physical characteristics and climbing ability of toads.