Localization in harsh and complex environments, such as industrial, confined or underground mines, is paramount. This paper presents a comparative localization study of wireless sensors in a complex environment, an underground mine, based on the received signal strength (RSS). Three estimation algorithms are tested to localize sensors in this harsh environment; the maximum likelihood estimation (ML), the two-step weighted least squares estimation (TWLS), and the generalized total least squares estimation (GTLS). The propagation model used in this work is obtained from conducted experimentation in a real mine environment. To the best of our knowledge, no comparative study was done between different localization algorithms in a mine environment. The aim of this work is to show the importance of the suitable choice of localization technique in difficult propagation conditions, allowing afterwards use and enhancing it depending on propagation parameters.