The pear psylla, Cacopsylla pyricola (Förster) (Hemiptera: Psyllidae), is a key pest of pear trees in Iran. One of the processes that regulate an insect population around the equilibrium position is the host plant effect. Therefore, the effect of plant variety on equilibrium level of pear psylla population was investigated for three pear varieties, including Shahmivea, Sebri and Coscia. A density-manipulation experiment was performed in a randomized complete block design with densities of 2, 4, 8, 16 and 32 pair of C. pyricola adults. The population models of Hassell and Ricker were fitted to the data and the minimum model was selected based on the model simplification. Populations recruitment curve were created and population equilibrium level of pear psylla was determined for each variety. The stability of the predicted equilibria was then determined. Long-term dynamics of pear psylla population was predicted for 100 generations by computer simulation of the best-fitted model. The results showed that the Ricker model could be fitted well to the data for all three varieties. The pear psylla population showed a density-dependence reaction. The equilibrium population values showed a difference of up to 36% for the three varieties. Predicted equilibrium was unstable for Shahmivea variety and while it was stable for two other varieties. In addition, computer simulation verified that long-term dynamics was fluctuated for Shahmivea variety and smooth for two other varieties. The effect of bottom-up forces of host plant on the magnitude and period of population fluctuations of pear psylla has been discussed.