The direct effect of radiative forcing induced by anthropogenic and naturally occurring aerosols is a major research subject. It also represents a source of uncertainty in the study of the global surface radiation budget. To better quantify aerosol forcing and its uncertainties, observational efforts have been launched to measure forcing at various geographical locations at different times throughout the year. This paper presents quantitative estimates of surface aerosol forcing based on ground-based observations during recent field campaigns (PRIDE 2000, SAFARI 2000 and ACE-ASIA 2001) and at the UCLA surface radiation measurement site. Measurements of transmitted solar radiance were obtained using narrow band spectral radiometers, including the Cimel CE-318 Sun-Photometer and Yankee Environmental System's Multiple Filter Shadow Band Radiometer (MFR-7). The measured radiances from the Cimel Sun-Photometer were processed by NASA's Aerosol Robotic Network (AERONET) office to produce aerosol optical depth (AOT). The Beer-Lambert Law was used to calculate the AOT's from the MFR-7's direct beam measurements after correcting for molecular scattering and ozone absorption. An array of pyranometers and pyrgeometers were also employed to measure the downward short wave and long wave fluxes respectively. We use methods developed by Haeffelin to estimate thermal offsets in the pyranometer data at the UCLA site and subtract them out for improved accuracy in the short wave data. Offsets were found to be no more than -5 Wm² . To reduce aerosol-forcing uncertainties under a cloudy sky, a cloud removal procedure adopted from previous investigators (Lin, Conant, and Ji) is also applied to the flux data. The surface aerosol radiative forcing (in Wm² ) at 0.498 µ m is then estimated by a differential technique for different air masses. Forcings for an air mass equal to 2 were found to be -92.1, -204.1 and -104.9 for PRIDE, SAFARI and ACE-ASIA respectively. These different values are attributed to the different types of aerosol particles studied during each field campaign. Additional methods developed by Conant and Ramanathan were also pursued in this study in order to refine the surface forcing estimates and the corresponding aerosol forcing efficiency. Lastly, in-situ measurements recorded during the field campaigns and at UCLA using aerosol samplers and sizers together with a detailed radiative transfer model are used to interpret the aerosol forcing estimates derived from observations.