Estimates of dynamic height anomalies from expendable bathythermograph (XBT) and TOPEX/Poseidon (T/P) sea surface height (SSH) measurements were compared along a transect at ~30 ¿S in the South Pacific. T/P SSH anomalies were calculated relative to a 5 year time mean. XBT dynamic height was calculated relative to 750 m using measured temperature and an objectively mapped climatological temperature-salinity relationship. The anomaly was obtained by subtracting out an objectively-mapped climatological dynamic height relative to 750 m. XBT temperature sections show evidence of a double-gyre structure, related to changes in shallow isopycnals near the gyre's center. XBT dynamic height and T/P SSH anomalies compare well with an RMS difference of 3.8 cm and a coherence above 0.7 for scales larger than 300 km. The differences between the two measures of dynamic height yield systematic patterns. Time-varying spatial averages of the differences are found to be related to changes in Sverdrup transport, zonal surface slope differences, and the 6 ¿C isotherm depth. Higher zonally averaged altimetry SSH than zonally averaged XBT height and larger northward transport from altimetry SSH than from XBT height correspond to gyre spinup determined from Sverdrup transport changes. This implies mass storage during gyre spinup due to the phase lag between the Ekman pumping and the full baroclinic Sverdrup response. Increases in the spatially averaged differences and zonal slope differences, associated with gyre spinup, correspond to shoaling in the 6 ¿C isotherm depth, requiring deep baroclinic changes out of phase with the 6 ¿C isotherm depth changes. ¿ 2000 American Geophysical Union