1. A model was set up to simulate the dynamics of a giant panda Ailuro
poda melanoleuca population, All parameters in the model were taken fr
om an existing panda population in the Shuidonggou Valley area of the
Qinling Mountains, China. 2. The model projects population dynamics an
d assesses population viability under changing conditions. 3. Small, f
ragmented populations face potential risks in terms of genetics and de
mographic processes. This study concerns the latter. 4. The simulation
indicated that: (i) the population is stable under present conditions
(but note that the model does not incorporate environmental stochasti
city); (ii) the environmental carrying capacity is the most important
factor restricting population increase; (iii) poaching would greatly a
ffect the probability of population survival; (iv) increasing the surv
ival of individuals under 1 year of age would increase population pers
istence substantially; (v) reducing subadult female mortality would ac
celerate population growth; (vi) shortening the interbirth interval wo
uld allow more rapid population growth, and its effect is greater than
changing other parameters, but may not be practicable; and (vii) immi
gration and reduction in young male mortality are of limited benefit t
o this small population but may be important in practical conservation
management. 5. The model suggests management strategies for giant pan
da conservation.