Oscillations in quasilinear stochastic hereditary systems and their application to specific problems are investigated. The algorithm of transition to a standard system with the help of the averaging method with small parameter in the presence of fast time is justified.In [1], the strong solution of a stochastic functional differential equation with Poisson perturbations (SFDEPP) is proved to be continuous with respect to a small parameter e > 0. This allows replacing the solutions of the SFDEPP with some Markov process [2][3][4]. As a consequence from the continuous dependence of the solution of the SFDEPP from parameter e > 0 , we will use the averaging method [5, 6] to establish the proximity of these processes on an asymptotically large time interval [ , / ] 0 T e . Moreover, we will present the method of reducing a second-order quasilinear SFDEPP to the standard form and prove the averaging theorem for this system. We will describe the splitting method, where the generating operator of the deterministic part of the stochastic equation on the chosen space C([ , ]) -t 0 of continuous functions is represented as a contour integral of a complex plane [7, 8]. On the probability basis ( , , , ) W F Ã ¹ [9], a random process R R +´® W 1 is defined as a strong solution [10-12] of the quasilinear SFDEPP dx t m x dt t dw t t u dt du f t dt t U ( ) ( ) ( ) ( ) ( , )~( , ) ( ) = + + + ò s l n (1)with the initial condition-® t 0 1 ; S is a Skorokhod space [13,14]; the random processes f R R : +´® W 1 and s : R R +´® W 1 and the random function l:Moreover, w t ( ) is a scalar Wiener process;~( , ): ( , ) ( , ) n n n t A t A t A = -¸is a centered Poisson measure [9, 10]; n( , ) ( ) t A t A = P ; and w t ( ) and~( , ) n t A are independent random processes. Denote by H t ( ) the solution of the deterministic equation [7, 15] dH t m H dt t ( ) ( ) = (5) with respect to the initial condition H t t t ( ) , , , . = = < ì í î 1 0 0 0 Following [7], we will call this solution the fundamental solution of (5).