Coffee crop is one of the main activities in the Amazon region, ranking as third in Peru. The purpose of this work was to evaluate the agronomic behavior of three coffee varieties (Catimor, Maragogipe and Castillo) established under an agroforestry system, using as shade trees Eucalyptus torreliana, Colubrina glandulosa and Pinus tecunumanii, whose experimental study was established in the district of Huambo, Province of Rodríguez de Mendoza, Amazonas-Peru during 2017. Treatments were distributed in a randomized complete block experimental design, with three blocks and nine treatments. The evaluated variables were plant height, height to the first leaf, number of leaves, general state of the plant and diameter of the stem. The methodology used consisted of individual analysis for each time, homogeneity test of variances and joint analysis over time for the same space, considering the interaction treatment*time, performing evaluations at 30, 60, 90, 120 and 150 days after crop seatlement. The results indicated that the variety of Castillo coffee either with shade of P. tecunumanii or with shade of C. glandulosa leads to a better agronomic behavior, and the variety of coffee Maragogipe both with shade of E. torreliana and with shadow of C. glandulosa shows a poor agronomic behavior.