1. Mantids visually orient to prey objects by means of rapid saccadic head movements which serve to image the target on the anterior regions of the large compound eyes. Small stationary flashes of light subtending 30′ of arc were used to elicit head movements. The topography of the saccade was studied by photographic and optoelectronic analysis. 2. High speed photography was utilized to study the temporal properties of individual saccades. The latency of the response ranged from a minimum of 0.15 s to a maximum of 1.0 s. Maximum angular velocities of 400 deg/s were achieved with large amplitude saccades. 3. Single stationary flashes produced saccades with latencies ranging from 0.30 to 1.0 s. When a saccade was not elicited by a single flash, a subsequent flash in an area 1.8° removed from the first would often evoke a response. The saccades produced by such multiple flashes yielded latency values ranging from 0.15 to 0.30 s. It, therefore, appears that a single stationary flash may potentiate the effects of subsequent stimulation. 4. The head orientation following a saccade is linearly related to the position of the stimulus. As previously described by Mittelstaedt (1957), the slope of the function relating head position to stimulus angle is less than unity. Slope values ranging from 0.65 to 0.83 were obtained in this study. 5. Orientation is dependent on the direction of the head movement. Large amplitude saccades directed laterally to a given stimulus location produce final head orientations which differ from those produced by saccades elicited to the same location but from the opposite direction. 6. For saccades proceeding in the same direction, final head orientation is independent of prior head orientation. 7. Visual feedback is unimportant in determining the final head position. Saccades elicited by stimuli whose duration was shorter than the mantid's reaction time produced the same final head angle as those saccades elicited by long duration stimuli. 8. After an initial large amplitude saccade was elicited, a small amplitude correction saccade often could be obtained by a second flash in the same location. Correction saccades were equally likely to occur following saccades elicited by long duration stimuli or short duration stimuli. Since reorienting saccades did occur following initial saccades produced by long duration stimuli, it must be concluded that visual feedback is not utilized to “home-inon” the target.