This study proposes a new instrument to measure cognitive load types related to user interface and demonstrates theoretical assumptions about different load types. In reconsidering established cognitive load theory, the inadequacies of the theory are criticized in terms of the adaption of learning efficiency score and distinction of cognitive load types. Since measurement of mental effort does not cover all types of cognitive load, a new way of isolating different loads is required. Previous studies have focused on designing interface to reduce extraneous cognitive load. However, interface may have the potential to enhance germane cognitive load because learners may construct their knowledge schemata with interface layouts. In this study, meaningful cognitive activities related to user interface were measured as germane cognitive load in web-based instruction. In addition, a usability test was used to gauge extraneous cognitive load. The results revealed that germane cognitive load and extraneous cognitive load are closely related, but other relationships were not significant. Germane and intrinsic cognitive load positively influences learning performance. Even though new instruments are still subjective and limited to interface in web-based instruction, the experiment strengthens cognitive load research. (Contains 4 tables and 4 figures.)