In this paper, we examine the following question: "do two conditions imply
the third one in the '2 out of 3' Lemma (see e.g. Lemma 9.5 in [Alp86, p. 6
8])?" We get a positive answer in two implications for p-nilpotent groups (
Section 2) and give some counterexamples in case of soluble groups (see Sec
tion 3).