Social media use is ubiquitous, especially among young adults. Negative consequences of social media use has been associated with engaging in upward social comparisons with others on social media. The current paper presents a series of two studies that seek to understand the nuances of social comparisons as they occur while browsing social media. In Study 1 (N = 161) we tested whether upward social comparisons would be less harmful when the comparer focuses on the similarities, rather than differences, with the comparison target. We observed a marginal interaction indicating that when thinking about similarities with the target, upward comparisons were slightly less detrimental to self-evaluations, than when thinking about differences. Study 2 (N = 320) tested other potential moderators of upward social comparisons suggested by traditional theories of social comparison to see if they would apply to a social media context. We predicted when a participant rates the comparison target to be more similar (vs. different) and distant (vs. close) and the domain to be more attainable (vs. less attainable), upward social comparisons will be less harmful to self-evaluations. Neither similarity nor closeness were significant moderators. Multilevel models showed a significant interaction between upward comparisons and attainability on self-evaluations (p < .001). Simple slopes indicated viewing the domain as attainable amplified the harmful effects of upward social comparisons on self-evaluations compared to viewing the domain as less attainable. These results are contrary to our predictions and suggest that traditional theories of social comparisons may not apply to a social media context. We also investigated how the target of the social comparison would impact the experience of the social comparison and its impact on self-evaluations. Results from Study 2 showed that participants who made comparisons with distant others (celebrities/strangers) were more likely to have higher upward social comparison scores, lower levels of similarity and closeness and view the domain as less attainable. These results display when we could expect harmful consequences following making a social comparison with a celebrity or stranger while viewing Instagram, posts. Study limitations and advantages are discussed along with future research to be explored.