Response contingent stimulation (RCS) is known to reduce stuttering in many cases. Time-out (TO) from speaking after stuttering is a particularly eOE ective form of RCS. However, the mechanism underlying the TO eOE ect is unknown. The demonstrated value of operant methods in the clinical control of stuttering in young children warrants renewed interest in exploring why pausing for a short time after stuttering would reduce the frequency of stuttering. In this study, selected linguistic measures were made for two school-age boys whose stuttering reduced during TO in a laboratory study. One boy showed a clear reduction in verbal output, accompanied by a reduction in lexical diversity, but the other boy showed no changes in linguistic measures. This ® nding is of interest because, in a previous acoustic analysis of the same speech samples, the ® rst boy showed no changes in acoustic speech measures of duration while the second boy did show changes in acoustic speech measures of duration. This raises the possibility that people who stutter may change their verbal output in response to TO, but in diOE erent ways. Although the present data do not establish a causal relationship between alterations in verbal output and reductions in stuttering, they raise the need for further research that might explain why TO might control stuttering.