An increasing number of people around the world engage in “crowdworking” (CW), defined as digital gainful work on intermediary internet platforms. Though this new work arrangement is a trilateral exchange involving the client, the crowdworker, and the platform, it still has an employment relationship at its core: A crowdworker is compensated for a personal service, and the work process is at least partly imposed on the crowdworker. This paper systematically reviews empirical research (k=118) on CW as an employment relationship, drawing on extant theoretical insights from human resource management and psychology. It uses an Input-Process-Output model and bibliometric network analysis to examine which topical areas of crowdworking-as-employment relationship have been covered, and which areas remain under-explored. The topical areas discussed in this paper comprise incentives, work design, the crowdworkers traits, skills, and working records as inputs; effort, affect, motivation, satisfaction, self-efficacy as process-level phenomena; and satisfaction, commitment, and performance as outputs. Reviewing these topical areas show that CW research focuses more on issues related to optimizing the task process from the platforms perspective rather than on topics of interest from the crowdworkers and the clients perspectives. The paper concludes by identifying five important but under-researched fields, namely long-term strategic workforce planning, legal issues, leadership styles, careers on platforms and employment relations. Therefore, it shows that compared to regular employment relationships, CW raises old problems in new and partly complex variations, on account of higher coordination efforts, fewer legal boundaries, crowdworkers paradoxical social roles, and intensive interactions with client and platform.