Plateau pikas are a keystone species and ecosystem engineers in alpine meadow ecosystems. A number of surveying methods have been used to estimate pika density, but the reliability of these methods is not known. In addition, better population density methodologies allow for more reliable density estimates. We therefore compared the relationship among several commonly used methods of estimating pika relative density and the absolute density (AD) of pika populations. This relationship was investigated in summer and winter pastures to determine whether distribution pattern of the pikas (patchy or uniform) would influence this relationship. During August of 2015 and August of 2016, we measured the relative and absolute population density of pikas in an alpine meadow in the County, Gansu province, northwestern China. Relative density was measured with three indirect and direct methods: the total burrow density (TBD), the active burrow density (ABD), and the direct counting density (DCD) at the peak of pika activity. AD was assessed by removal sampling. Our results showed that the relative population density with all three survey methods was significantly related to the absolute population density. In particular, DCD at the peak of the pika activity showed the best correlation with AD. A simple linear model showed the effect of grazing time by livestock had a significant effect on TBD and ABD. These results imply that using DCD as a method to survey at the peak of the pika activity is more reliable than other methods.