[Context and motivation] With the advent of agile development, requirements are increasingly stored and managed within issue tracking systems (ITSs). These systems provide a single point of access to the product backlog, bugs, ideas, and also tasks for the development team to complete.

[Question/problem] ITSs combine two perspectives: representing requirements knowledge and allocating work items to team members. We tackle a knowledge problem, addressing questions such as: How are requirements formulated in ITSs? Which types of requirements are represented? At which granularity level? We also explore whether a distinction exists between open source projects and proprietary ones.

[Principal ideas/results] Through quantitative content analysis, we analyze 1,636 product backlog items sampled from fourteen projects. Among the main findings, we learned that the labeling of backlog items is largely inconsistent, and that user-oriented functional requirements are the prevalent category. We also find that a single backlog item can contain multiple requirements with different levels of granularity.

[Contribution] We reveal knowledge and patterns about requirements documentation in ITSs. These outcomes can be used to gain a better empirical understanding of Agile RE, and as a basis for the development of automated tools that identify and analyze requirements in product backlogs.