Context. Asteroid-pair age estimations are usually performed by backward integration of possible orbits of pair components, taking planetary perturbations into account along with the Yarkovsky effect. It is assumed that uncertainties coming from the backward integration process itself are small, so to reduce uncertainties in estimations, effort usually focuses on reducing uncertainties in the initial orbital elements and the dimensions of the pair components. Aims. This work aims to evaluate the role of the frequently ignored nonplanetary perturbers in asteroid-pair age estimation. When their role is not negligible, the ages of the youngest known pairs can be roughly re-estimated. Methods. The orbital evolution of several asteroid-pair components and the close approaches between components during the last ≈43 kyr are investigated. Initially, the force equations consisted of only planetary perturbers. The three largest main belt perturbers were added afterwards. Finally, as many as 262 massive main belt perturbers were included. The effect of main belters on age estimation is assessed by comparing the dates of the closest approaches between pair components. The range of the Yarkovsky effect is simulated very roughly, only for comparison purposes. Martes -2005 WW 113 , which was previously reported as very young, had slightly higher encounter velocities and longer nominal minimum distances, calling into question whether they could have formed within the interval studied or in some previous mutual encounter of pair components. Conclusions. The effect of main belt perturbers on asteroid-pair age estimation is buried in the huge age uncertainties caused by the Yarkovsky effect until the rotation direction of pair components is recognized.