This study aimed to investigate the relationships between college students' behavioral and cognitive engagements while performing an online collective information searching (CIS) activity. The activity aimed to assist the students in utilizing a social bookmarking application to exploit the Internet in a collective manner. A group of 101 college students in Taiwan participated in the research procedure, and performed the CIS activity to glean quality online resources for the given search assignment. The actions taken and annotations and comments made during the activity were recorded as log data, and used as the main resource for later analyses of behavioral and cognitive engagements in the activity. Through cluster analysis of the students' contributions to the CIS activity, four categories of behavioral engagement were identified, namely "Hitchhiker," "Individualist," "Active" and "Commentator," to represent the students' investments in performing the activity. Furthermore, to explore the students' cognitive engagement in the activity, content analysis of the verbal transcripts of their annotations and comments was conducted based on the refined coding framework of the present study. The results of further cluster analysis revealed that the students' cognitive engagement levels could be identified as "Deep" and "Surface." Through comparison of their behavioral and cognitive engagements, the findings revealed that the students with "Active" behavioral engagement tended to exhibit a "Deep" level of cognitive engagement. It is therefore suggested that both behavioral and cognitive engagements are critical to participatory learning with practice in CIS activities.