Three range-wide surveys have been conducted, in the mid-1970s, mid-late1980s, and 2000–2002. All surveys were based on incidence of sign, but techniques varied, so results are not directly comparable. Present best estimates indicate a total wild population between 1,000–2,000. Greater protection of forests and from poaching in recent years suggest that panda populations should be increasing, but this has not been confirmed empirically.

Results from the most recent survey, coordinated by the State Forestry Administration (SFA) of China and World Wildlife Fund (WWF), indicated a total population of ~1600 individuals. This is over 40% higher than previous estimates. It is believed that the increase in the estimated number of pandas is due largely to differences in survey methodology and a larger search area, as well as possibly an actual increase in panda population size in some areas. Conversely, in other areas, habitat conditions were deemed to be worse and panda numbers lower in 2000–2002 than in the 1980s survey.

The most recent population estimate was based on differentiating individual pandas from measurements of bamboo fragments in scats. It is known that different age classes of pandas have different bite-sizes of bamboo (Schaller et al. 1985), but the validity of differentiating individuals of the same age class based on bite-sizes has not been well tested. Recent information on DNA-identified scats suggests that the bite-size method may underestimate population size in some cases (e.g., dense populations; Zhan et al. 2006).

Many surviving wild giant panda subpopulations have fewer than 50 individuals (Loucks et al. 2001). No major reductions in the genetic diversity of these populations is apparent, although they likely experienced modest genetic losses from a much larger ancestral population (Lü 2001). Some controversial research suggests that the Qinling (Shaanxi Province) population is a genetically isolated and distinct subspecies (Wan et al. 2005)

Decreasing