The purpose of this study was to explore the effectiveness of using learner-generated and instructor-provided multimedia annotations on foreign language reading comprehension and attitudes. The four research questions are: (1) what are the effects of using different multimedia annotations on reading comprehension for learners of different cognitive learning styles (field-dependent and field-independent)? (2) What are the effects of using different multimedia annotations on reading comprehension for learners with different learning abilities (higher-level and lower-level)? And (3) what are learners' attitudes toward using a multimedia annotation system? The results of this study are listed as follows: First, for reading comprehension, the learner-generated annotation group performed better than the instructor-provided group, no matter which cognitive learning style they were. Second, higher-level learners with learner-generated annotation performed better than those with instructor-provided annotation. However, the difference between lower-level learners with learner-generated annotation and those with instructor-provided annotation was not significant. Finally, learners had positive attitudes toward multimedia annotation use and thought text annotation was the most useful of the different types. (Contains 11 tables and 1 figure.)