The reflex regulation of reticulo-ruminal movements was analysed in the present experiments using an electrophysiological 'single fibre* recording technique. The afferent input to and the efferent output from the 'gastric (retioulo-ruminal) centres' was repeatedly sampled by recording from single gastric units dissected from the cervical region of the left vagus in 8 sheep anaesthetized with chloralose and 70 with halothane. Primary cycle movements of the reticulo-rumen were usually evoked by distending a reticular balloon with 400-600 ml air. By recording from afferent gastric units, it was found that the majority of gastric mechanoreceptors are slowly-adapting 'in series' tension receptors situated in the muscle layers of, principally, the reticulum, the reticuloruminal fold and the cranial sac (dorsal rumen). The afferent discharge from these receptors increases during passive distension and during isometricallyrecorded contractions. Receptors in the lips and the floor of the reticular groove and oma sal canal respond both during a contraction and, particularly, to pressure. The 'resting discharge' generated by tension receptors is largely determined by the intrinsic motility of smooth muscle cells. The mean conduction velocity in afferent gastric fibres is 12.4 m/sec. By recording from efferent gastric units, at least 7 distinctive types were discernible and, by relating their discharges temporally to movements of the reticulum and of the rumen, it wa3 concluded that Types I, II and III occur in fibres innervating the reticulum or associated structures, Type IV the rumen and Types V, VI and VII other gastric structures not yet identified. In only Type VII units is there a 'resting discharge' during the quiescent part of the gastric cycle. By cold blocking vagal nerves, it was demonstrated that separate efferent fibres innervate the reticulum and the rumen and, that the dorsal vagal trunk carries predominantly excitatory afferent fibres to the gastric centres whereas the ventral vagal trunk carries predominantly either excitatory or inhibitory fibres depending on the experimental conditions. By recording from afferent and from efferent gastric units at the same time as altering conditions in the reticulum either physically or with certain drugs, the changes in the afferent input to and the efferent output from the gastric centres evoked by these manoeuvres were recorded and estimates of the total and the central reflex time for gastric reflexes were made. It is concluded from the present investigation that: (a) the co-ordination of the complex sequence of primary cycle movements is a function of the 'gastric centres', through their ability to determine the parameters and temporal interrelationships of efferent nervous discharges in the various types of gastric units innervating different regions of the forestomach. (b) the tonic afferent input from 'in series' reticular tension receptors during the quiescent period of the primary cycle provides a reflex •drive* to the gastric centres and largely determines the rate, the duration and the amplitude of reticular and ruminal contractions. (c) the enhanced afferent input to the 'gastric centres', occurring during a reticular contraction recorded under isometric conditions, modifies the form, the amplitude, the duration and the delay in onset of the later parts of the contraction sequence of the reticulum and the rumen.