Observations during a field experiment in Oklahoma in the fall of 1995 have been analyzed to show the influence of clouds and water vapor on broadband absorption of solar radiation in the atmosphere. The fraction of solar flux incident at the top of the atmosphere absorbed during 34 days was 0.24¿0.04. For a 7-day subset during which skies were clear, it was 0.22¿0.02. Both clouds and water vapor contribute to day-to-day variability. The water vapor effect is systematic, increasing absorption by 0.04--0.05 per g cm-2 increase of column water vapor. Clouds, on the other hand, increase or decrease total column absorption, depending on cloud configuration. Most of the average cloud effect is due to water vapor, which was highly correlated with cloud amount during the experiment. Overall, model calculations underestimate atmospheric absorption by 0.06, with 25--55% of the discrepancy associated with water vapor, ~15% with clouds, and the remaining 30--60% independent of either variable. ¿ 1999 American Geophysical Union