Recognition of a self‐image in a mirror is investigated using the mark test during which a mark is placed onto a point on the body that is not directly visible, and the presence or absence of self‐directed behaviors is evaluated for the mirror‐observing subjects. Great apes, dolphins, possibly elephants, and magpies have all passed the mark test, that is, displayed self‐directed behaviors, whereas monkeys, crows, and other animals have failed the test even though they were able to use a mirror to find a not‐directly‐visible object. Self‐directed behavior and mirror use are prerequisites of a successful mark test, and the absence of these behaviors may lead to false negative results. Epstein, Lanza, and Skinner (1981) reported self‐directed behavior of pigeons in front of a mirror after explicit training of self‐directed pecking and of pecking an object with the aid of a mirror, but certain other researchers could not confirm the results. The aim of the present study was to conduct the mark test with two pigeons that had received extensive training of the prerequisite behaviors. Crucial points of the training were identical topography (pecking) and the same reinforcement (food) in the prerequisite behaviors as well as sufficient training of these behaviors. After training for the prerequisite behaviors, both pigeons spontaneously integrated the learned self‐directed and mirror‐use behavior and displayed self‐directed behavior in a mark test. This indicates that pigeons display mirror self‐recognition after training of suitable ontogenetic contingency.