The balances of momentum and heat were studied for an 18-day period in October--November 1983 during the Mixed-Layer Dynamics Experiment at a site 650 km of central California. Data wer collected from a drifting platform from which surface winds, air temperature, and ocean currents (from a string of vector-measuring current meters) were measured. The position of the unattended drifter was determined by LORAN-C on the plateform. Currents in the surface layers were partially wind driven and partially coherent with low-frequency currents at deeper levels. The latter currents, though coherent, were sheared, turned significantly with depth, and were presumably nearly geostrophically balanced. Vertically differencing velocity in the mixed layer improved the balance between wind and current by removing a portion of the geostophic flow. The complex correlation coefficient between wind stress and current integrated to 38 m was 0.76 for current relative to 38 m yet only 0.36 for the undifferenced case. Surface heat flux magnitudes were not clearly reflected in the heat content of the near-surface layers although variability was. Over the first 13 days of the experiment, the correlation coefficient between daily-averaged heat flux and heat content change in the top 38 m was only 10 W m-2. Heat balance must have been restored by advection. Vertical advection was rejected as an explanation because heat content changes in layers defined by isotherm depths showed similar imbalances with regard to surface heating. Horizontal advection was estimated by two methods: (1) by using sparse thermistor chain tows around the current meter drifter to construct a history of horizontal temperature gradients and (2) by using current profiles at moderate depths to estimate pressure gradients (essentially through geostrophy), hence buoyancy gradients (by the hydrostatic relation), hence temperature gradients. The two methods gave very different estimates, though both tended to redress the imbalance between surface heat flux and local heat storage.