This study examined the effects of response time and message content on the growth patterns of discussion threads in computer-supported collaborative argumentation. Event sequence analysis was used to measure response times between threaded messages and responses containing arguments, evidence, critiques, evaluations, and other comments from online debates. The results supported and contradicted the findings of Hewitt and Teplovs (1999). Response rates overall declined at a rate of 17% per day in wait time across all message categories. On the other hand, the posting of critiques and particular types of argumentative exchanges produced higher response rates of 0.72 and higher, and their average wait times of 1.04 days were significantly longer than those of other message types. The debate format and use of message labels may have produced sufficient argumentative exchanges to produce high response rates despite the long response times, which in turn helped sustain the growth of discussion threads. (Contains 1 figure and 4 tables.)