The new version 9 of Wolfram’s Mathematica includes a set of quite diverse social network analysis functions. They include a range of import formats (including Pajek, GXL and GraphML) to load your own data, but you can also extract data from social media platforms like Twitter and Facebook. Networks can be visualised and different lay-out algorithms can be used. It also includes a range of analysis techniques like communities, centrality, cliques and homophily. In addition, scale-free networks can be simulated.
I have not yet tried the software myself, so a full review is still pending. At first sight the combination of these functions makes Mathematica 9 look quite comprehensive for SNA. But there is no indication that its analytical capabilities are as exhaustive as Pajek or that its import/export formats are as flexible as UCINET. Other alternatives with a nice graphical user interface like Cytoscape and Gephi also include most of the features of Mathematica 9, although maybe the latter works faster than the other two? I believe the biggest strength might be that these SNA tools are integrated in a large mathematical program with diverse functionality, which might allow for pushing SNA results further.
Feel free to comment if you have experience with Mathematica. Would love to hear about how good this is.