Incompleteness in fossil specimens is the main obstacle that has to be overcome with the identification of fossils and the analyses using them. Landmark analysis is a tool that can be used to understand the variations in different organismal factors based on morphology with adequate preparation and properly aligned photographs of specimens. The goal is to apply these methods to assess generic level identification of Miocene Astriclypeidae, including Astriclypeus and Echinodiscus, based on incomplete onsite specimens from the Yehliu Geopark, Taiwan; a locality in where the removal of fossil specimens is not permitted. Two datasets were chosen utilizing the available fossil specimens from the site: a three-point dataset and a seven-point dataset. Linear measurements of onsite specimens were also recorded for comparison. Results from Principal Component Analysis (PCA) showed that samples of Astriclypeus and Echinodiscus formed distinct clusters based on three-point dataset. A similar trend with distinct clusters for the two genera was also evident with the seven-point dataset. Furthermore, the effectiveness of this method is reinforced by the results of the independent study using the traditional method with linear measurements. Geometric morphometrics can specifically identify where morphological variation occurs and is concentrated based on the chosen landmark arrays, and such morphological variations cannot be detected easily with the linear and angle measurements alone. This study shows that the landmark analysis can be used efficiently for a generic level identification based on incomplete specimens.