People experiencing cancer use the Internet for many reasons, particularly for social support. This study sought to determine how social support content within online support communities for different cancers varied according to cancer survival rate. A quantitative content analysis was conducted on 3717 posts from eight online communities focused on cancers with high and low 5-year relative survival rates. Using Optimal Matching Theory, we predicted that low-survival-rate communities would have more emotional support content than high-survival-rate communities, and high-survival-rate communities would have more informational support content than low-survival-rate communities. Emotional support content was consistently more common than informational support. Overall, high-survival-rate communities had a greater proportion of posts containing emotional support content (75%) than low-survival-rate communities (66%) (χ21 = 20.89 [n = 2235], P < .001). Furthermore, low-survival-rate communities had a greater proportion of posts containing informational support content (46%) than high-survival-rate communities (36%) (χ21 = 21.13 [n = 2235], P< .001). Although the relationships between survival rate and support types were significant, they were not as hypothesized. Deviations from theoretically predicted results suggest that individuals experiencing low-survival-rate cancers may have a greater desire for informational support online than individuals experiencing high-survival-rate cancers.