This study explored tolerance and intolerance in political discourse on Twitter during the U.S. 2016 election. It was a combination of social network analysis of four Twitter networks the day before the election, the day of the election, the day after the election, and four days after the election (November 7, 8, 9, 12) and content analysis of 1,114 tweets from 40 largest clusters of the four networks. The study found significant differences between content frames (tolerant, intolerant, and neither) in relation to out-degree centrality. However, there were no significant differences between tolerant and intolerant content in relation to in-degree centrality, betweenness centrality, and closeness centrality. Findings show similarities among the overall network structures across four days as all of the networks had low centralization (no hierarchical structure), low density, high modularity, and low reciprocity scores. The networks were not polarized; instead, they were divided into several small clusters with mixed conversations about both in-group and out-group candidates. This finding is in contrast with previous studies that found political discourse in online social networks is highly polarized (Adamic & Glance, 2005; Cha et al., 2010; Himelboim et al., 2013; Himelboim et al., 2017; Smith et al., 2014). Another major finding of this study was that Twitter users were more intolerant than tolerant and the percentage of intolerant tweets was doubled the day after and four days after the election.