Temperature and humidity fluctuation data were collected on the R/V Flip during BOMEK (Barbados Oceanographic and Meteorological Experiment) in May 1969 and during a pre-BOMEX trial cruise near San Diego in February 1969. The program was a cooperative one with a number of groups on board. In a companion paper with the personnel from the University of British Columbia, the equipment and data analysis are described in more detail and the results of measurements of the fluxes of momentum, moisture and sensible heat are given. The emphasis of this paper is on the similarities and differences between temperature and humidity and on their cospectra and correlations with the velocity fluctuations. During the San Diego experiment it was found that humidity and temperature fluctuations were very similar with only minor differences. During BOMEX the humidity fluctuations were similar to those obtained during the San Diego experiment and to measurements over land, but the low-frequency ends of the temperature spectra are much lower in BOMEX than in San Diego and, consequently, so are the low-frequency ends of the temperature-vertical velocity cospectra. On the basis of our results and comparison with other results we conclude that the humidity spectra and humidity, vertical velocity cospectra have “universal forms” when normalized according to the Monin-Obukhov similarity theory. The temperature spectra cannot be interpreted in terms of similarity theory because this theory does not include the effects of radiative transfer. These radiative transfer effects depend in part on the moisture content of the air; when it is high, the temperature field, particularly gradients and large-scale fluctuations, is strongly affected. Having identified such effects for BOMEX, similar but much smaller effects can be seen in the San Diego results. Thus, temperature and humidity are not similar scalars and measurements of one are not necessarily indicative of the behavior of the other.