Ever since the term ‘CNN effect’ was coined to describe the impact of a 24-hour news cycle, it has been clear that public opinion was subject to constant potential influence and flux. Unfortunately, survey research, the traditional method of charting public sentiment, is slow. It is also expensive. And survey measures of public opinion are inevitably constrained by questionnaires, with little room for spontaneous expressions of opinion. Crucially, none of these limitations apply to the measurement of public opinion via social media ‘big data’. Measurement is continuous and instantaneous, providing a live monitor of public mood that can be tracked through a presidential debate, an election campaign, a foreign policy crisis, or a full parliamentary term. Respondents need not be persuaded or paid but participate voluntarily. And data from Twitter, Facebook, and so on are genuinely observational and non-reactive rather than manufactured, capturing the shades and flavours of public opinion. However, while the advent of the Internet (and especially Web 2.0) has occupied scholars studying party campaigning, elite communication and citizen participation, only recently has attention turned to social media as an outlet for the expression – and hence the measurement – of public opinion. This workshop is about understanding that source of data, exploiting it as far as possible, and recognising its limitations. We welcome three broad types of papers: empirical studies of public mood via social media; methodological work generating and validating social media measures of opinion; and theoretical contributions on the nature of public opinion measured via social media or on the role of these forums in processes of democratic deliberation and representation.