In a wireless network (WN), the wireless devices generally localize themselves with the help of anchors that are pre-deployed in the network. Some of the techniques commonly used for localization are Time of Arrival (ToA), Time Difference of Arrival (TDoA), Angle of Arrival (AoA), and Time of Flight (ToF). In the wireless domain, measurements are susceptible to errors resulting from the nature of the medium, the relatively low precision, and the presence of obstacles, which produce Non-Line Of Sight (NLOS) errors. The NLOS errors are a major concern as they could result in significant degradation in accuracy. In this paper, we propose an efficient technique that uses the distance estimates of the device from a group of anchors to localize the device with better accuracy in the presence of NLOS errors. Our technique is based on the notion that in general, for any estimate, the proportion of the NLOS error can be upper bounded. Using this upper bound information our technique reduces the uncertainty in the position of the wireless device that is being localized. The technique is distributed and is simple. In comparison to the standard localization procedure, where localization is done independent of the presence of NLOS errors, our technique uses the information about the NLOS error bounds to improve the accuracy of estimation. Simulation results show that our technique reduces the position error of the wireless device by 40% on an average and by at least 80% in the best case. The uncertainty in localization is also reduced significantly.