During the last few years at the Hampstead Clinic we have studied a small number of children who had at the time of their diagnostic assessment or previously had an imaginary companion. In no case was the imaginary companion the cause for referral. Usually, its existence was elicited more or less accidentally during the course of the diagnostic investigation.

When we explored this fantasy further, we were surprised to learn that only rarely did the imaginary companion play a significant role in the analysis of these children. In fact, we know of only two children who directly and frequently referred to their imaginary companions during the analytic sessions. One such case was described in the literature (O. Sperling, 1954). Since our clinical material is limited, we can only state what our experience was,