Internet Protocol (IP) cameras have become virtually omnipresent for organizations, businesses, and personal users across the world, for the purposes of providing physical security, increasing safety, and preventing crime. However, recent studies suggest that IP cameras contain less than ideal security and could be easily exploited by miscreants to infringe user privacy and cause even bigger threats. In this study, we focus on the IP cameras without any password protection. We conduct a large-scale empirical investigation of such IP cameras based on insecam.org, an online directory of IP cameras, which claims to be the largest one in the world. To this end, we have monitored the site and studied its dynamics with daily data collection over a continuous period of 18 days. We compute daily number of active IP cameras and new cameras on the site, and infer people’s usage habit of IP cameras. In addition, we perform a comprehensive characteristic analysis of IP cameras in terms of the most used TCP/UDP ports, manufactures, installation location, ISPs, and countries. Furthermore, we explore other possibly existing security issues with those cameras in addition to no password protection. We utilize an IP scanning tool to discover the hidden hosts and services on the internal network where a vulnerable IP camera is located, and then perform a vulnerability analysis. We believe our findings can provide valuable knowledge of the threat landscape that IP cameras are exposed to.