In recent years, companies' responsibilities towards society have expanded significantly, and former business models have been revised in line with a more pluralist approach, which takes stakeholders, sustainability, business ethics and transparency into account. These considerations are increasingly reflected in corporate sustainability reporting, i.e., statements reflecting the impact of business practices related to economic, environmental and social perspectives. However, the data contained in such documents are reported independently of the financial aspects, implying a disconnection between the three pillars of business strategy. In order to avoid this shortcoming, some leading companies have developed integrated reporting, which expresses the interconnections between a firm's strategy, governance, performance and prospects, as well as the contexts within which it operates. The aim of this paper is to examine the influence of one of the most important institutional factors, the legal system, on the development of integrated reports. The analysis of a non-balanced sample of 750 international companies for the years 2008–2010, in which a logit methodology is applied to panel data, reveals the impact of this institutional factor. The results show that companies located in civil law countries, and where indices of law and order are high, are more likely to create and publish a broad range of integrated reports, thus favouring decision-taking by the different stakeholders. On the basis of these findings, two policy recommendations are proposed: first, it may be necessary to establish national laws and protection mechanisms to promote and ensure holistic transparency. Second, managers must be able to decide on the appropriate disclosure practices in the context of their own legal environment in order to obtain maximum benefits from their decisions. The results should be considered as having only preliminary value due to the limited number of firms that disclose this typology of reporting, and because of the limitations of the data sample.