Simultaneously measured Eulerian currents and spatially extensive subsurface temperatures have provided a time series of eigth synoptic, three-dimensional views of the Gulf Stream frontal zone along the Carolina continental margin. Two large-amplitude meanders were observed to progress through the study area between Charleston and Cape Hatteras during February 1979. Each meander had a vertically coherent, skewed wave-like subsurface structure. The Eulerian velocity and temperature signatures produced by the meanders at the 250-m level over the 390-m isobath reflect this skewness. At a particular instrument, the in-phase increases in temperature and downstream velocity associated with an approaching meander crest occurred during a longer time interval than did the more rapid decreases in these quantities following the crest's passage. Typically, the downstream velocity component at this level fluctuated from about -20 cm s-1 to near 100 cm s-1, while the cross-stream component varied approximately ¿25 cm s-1 about a near-zero mean. For a particular meander, the maximum in the offshore velocity component led the downstream maximum in time in a manner typical of progressive wave motions; however, the lead time was always less than one quarter of a meander period implying that u and v were not in quadrature, as would have been the case for stable waves. The two meanders were observed downstream of the area off Charleston where a seaward deflection of the stream is often found. Subsurface temperature data from February 10, 1979, show that on that date the degree of deflection was greatest near the surface, and that almost no deflection existed within the deeper reaches of the water column. According to later data, the deflection decreased as the meanders progressed alongshore away from the area, suggesting that the vertical structure of the deflection observed on the tenth may have been associated with the late stages of a meander passage. Filament of warm Gulf stream water extended southwestward ''behind'' the crests of the two meanders. The filaments were relatively shallow features, extending from the surface to a depth of a few tens of meters. They were oriented essentially parallel to the bottom contours over the outer shelf and upper slope, and were separated from the main body of the Gulf Stream by cool water. The presence of the cool water between the stream and the filaments at the surface was due to upwelling of water from deep within or below the main stream. Peaks in the time series of vorticity components indicate that maximum cyclonic relative vorticity occurred behind the meander crests, in the leading portion of the trough near where a warm filament joined a meander crest. The meanders may have been initiated upstream of our studyarea, and then ''amplified'' by the deflection process off Charleston Energy flux calculations for the region off Onslow Bay indicate that meander kinetic energy was being coverted to mean energy there. It seems likely that the deflection produces meander growth within the 100 km or so immediately downstream of Charleston, to subsequently have the meander energy reconverted to mean energy farther downstream.