The absolute bias calibration for the GEOS 3 intensive mode altimeter has been measured by using two satellite passes whose ground tracks were within 1 km of the Bermuda laser station. The Bermuda laser tracked on the two passes-rev 4553 on February 25, 1976, and rev 5471 on April 30, 1976-and was supported by two other NASA lasers on one pass and by the NASA Spacecraft Tracking and Data Network on the other pass. For each pass the altimeter data around Bermuda was smoothed and extrapolated to the point closest to overhead at the laser site. This point was used for calibration, eliminating almost entirely the effects of geoid model error on the resulting altimeter bias estimate. After correcting for tide heights and sea state effects, the two passes give calibration biases which are in agreement to within 3 cm and have a weighted mean of -5.61¿0.16 m for correcting altimeter measurements to the center of mass of the spacecraft (i.e., including the antenna tracking point correction). For data correction, this bias must be subtracted from the measured values. Since a sea state bias correction has been used in the bias estimation, a different bias is more appropriate data users not employing a sea state bias correction. For such users, a bias of -5.51 m, appropriate for moderate seas (H1/3? = 2 m), is recommended. It was found impossible to reconcile the two calibration passes, as well as a set of altimeter crossovers in the middle of the GEOS 3 calibration area, without allowing for a data time tag error. On the basis of a selected set of four crossovers and an assessment of probable sources of timing error, it was concluded that one interpulse period (10.24 ms) should be added to the data time tags. This time tag correction should be used with the above bias value.