Radiochromic film (RCF) is attractive as a thin, high resolution, 2D planar dosimeter. We have studied the uniformity, linearity, and reproducibility of a commercially supplied RCF system (model MD-55). Forty 12 cm long strips of RCF were exposed to uniform doses of 6 MV x rays. Optical density (OD) distributions were measured by a helium-neon scanning laser (633 nm) 2D densitometer and also with a manual densitometer. All film strips showed 8%–15% variations in OD values independent of densitometry technique which are evidently due to nonuniform dispersal of the sensor medium. A double exposure technique was developed to solve this problem. The film is first exposed to a uniform beam, which defines a pixel-by-pixel nonuniformity correction matrix. The film is then exposed to the unknown dose distribution, rescanned, and the net OD at each pixel corrected for nonuniformity. The double exposure technique reduces OD/unit dose variation to a 2%–5% random fluctuation. RCF response was found to deviate significantly from linearity at low doses (40% change in net OD/Gy from 1 to 30 Gy); a finding not previously reported. To study the tradeoff between statistical noise and spatial resolution, OD was averaged over blocks of adjacent 50 μm pixels (ranging from 1×1 to 10×10 pixels). Reproducibility, defined as the standard deviation of repeated single-pixel measurements on separate film pieces, was 2% at 30 Gy for a resolution of 0.25 mm. With careful correction for nonlinearity and nonuniformity, RCF is a promising quantitative 2D dosimeter for radiation oncology applications.