This study examined the following issues: (a) If vocabulary effects are minimized, will there be SES differences in the use of conditional forms by Caucasian children who have reached the fifth grade? (b) Even if SES differences occur in the use of recognizable conditional forms, will children within the same SES be able to transmit conditional information to one another? (c) If SES differences are found in transmission of conditional information, are they attributable to lack of comprehension of the conditional logic, or to difficulty in expressing the logic? One hundred and thirty-two fifth grade Caucasian children, who were divided into lower and middle SES groups of approximately equal IQ. Dyads were formed by combining children into the four possible speaker-listener combinations. The task required speakers to use a conditional communication in describing the rules of a game to a listener. Results showed that, as listeners lower and middle SES children responded equally well to the communications of both lower and middle SES speakers. However, as communicators, the lower SES children performed more poorly than the middle SES: All listeners had equally great difficulty following the instructions provided by the lower SES communicators. The data suggests that SES difference can not be attributed to lack of comprehension of the conditional logic. Instead, it appears that the lower SES children were difficult to understand because of strong tendencies toward egocentric communicational patterns. (Author/JM)