We recorded the blows of gray whales during their southbound migration past central California in January 1994, 1995, and 1996, using thermal imaging sensors. For our sampling purposes, we defined day (0730–1630) and night (1630–0730) to coincide with the on/off effort periods of the visual counts being conducted concurrently. We pooled data across the three years of sampling and tested for diel variation in surfacing interval, pod size, offshore distance, migration rate, and swimming speed by comparing paired day/night means for samples collected within the respective 24-h period. We performed these tests using data from the entire migration period and then repeated the tests for samples collected prior to and after the approximate median migration date (15 January). Over the entire migration period we observed larger diurnal pod sizes (x̄day= 1.75 ± 0.280, x̄night= 1.63 ± 0.232) and greater diurnal offshore distances (x̄day= 2.30 ± 0.328 km, x̄night= 2.03 ± 0.356 km) but found no diel variation in surfacing interval. For the entire migration period, the nocturnal migration rate (average number of whales passing per hour) was higher than the diurnal rate. During the first half of the migration we detected no diel variation in pod size or surfacing interval, but diurnal offshore distances were larger than at night (x̄day= 2.28 ± 0.273 km, x̄night= 1.96 ± 0.318 km). Diurnal and nocturnal migration rates prior to 15 January were not different. During the second half of the migration, there was no diel variation in surfacing interval, pod size, or distance offshore, but the nocturnal migration rate was higher (28%, SE = 11.6%) than the diurnal rate. We found no diel variation in swimming speed in any comparison. We propose that later migrants socialize more during the day, which effectively slows their diurnal rate of migration relative to nocturnal rates.