Clouds, along with column water vapor, are the principal control of the surface radiation budget. Surface-based pyranometers and pyrgeometers provide measurements of the total flux but do not allow us to discriminate the effect of clouds on fluxes. We previously developed a method to compute surface cloud effect for shortwave radiation by combining time series measurements of total and diffuse shortwave measurements. More recently, we have developed an analogous method for longwave radiation. Based on these approaches, we then developed a technique to estimate daytime cloud amount from shortwave flux measurements and now have implemented as well the LW effective sky cover technique of Durr and Philipona (JGR, 2004). These cloud amount estimates are accurate to +/- 0.1 in general. We have applied these methods to the multi-year radiation measurements at the ARM sites in Oklahoma, Barrow Alaska, Nauru, and Manus Island PNG. The results presented here include seasonal and interannual variability and site-to-site comparison of cloud amount and surface cloud effect. We also include a brief comparison of day/night LW cloud amount and a comparison of daytime cloud fraction estimated from both SW and LW. These data sets provide a critical constraint on the ability of forecast and climate models to simulate adequately both cloud amount and the radiative effect of clouds on the surface radiation balance and, therefore, on the transfer of heat from surface to atmosphere. The techniques we describe can be applied to any well-calibrated broad-band radiometer suite such as those at the Baseline Surface Radiation Network (BSRN) sites. Such measurements and derived quantities should be included in any global climate observing system of the future.