Pregnant women, carrying end of gestation fetuses (38-41 weeks gestational age) were either gently rocked in an anterior posterior plane in a rocking chair or gently swayed laterally in a garden glider. Stimulation lasted for either 5 or 26 s (R5 & R26, SW5 & SW26). Fetal heart rate (HR) (bpm) and ultrasonically detected fetal movements were recorded 26 s before and during stimulation. Heart rate data were processed two ways: (a) ANOVAs performed on average values for each group and (b) comparisons of the proportion of subjects in each group showing HR accelerative change (AC), HR decelerative change (DC), or no change (NC) using chi2 statistics. HR direction and amplitude of changes were determined with the help of a conservative procedure previously developed and fitted to multidirectional HR responses (Lecanuet, Granier-Deferre, Jacquet, & Busnel, 1992). Results of ANOVAs indicated that a 26-s rocking stimulation induced a significant average HR acceleration (4.53 bpm) compared to a nonstimulated (control) group. Five seconds of rocking had a weaker, but still significant, effect on HR acceleration, but the two swaying conditions did not reliably affect either the HR or behavioral measures. Processing data nonparametrically confirmed that R26 stimulation evoked mostly AC (72%) and identified a gradual decrement of AC and DC percentage responses from Group R5, SW26 to Group SW5 while NC percentages raised. Hypotheses concerning the target sense activated by this stimulation were considered in relation to responses to auditory stimuli.