Projected index computed tomography (PICT) is a newly developed technique that uses the measured optical path length from multiple optical coherence tomography (OCT) scans through a semitransparent sample to computationally reconstruct tomographic images based on the spatial variations of the refractive index. Since the index of refraction of most samples is not constant, a depth-wise distortion due to the varying index of the sample is evident. Using a highly reflective reference surface placed behind the sample, optical path length measurements yield an aggregate index value for each beampath through the sample. Rotating the sample allows a data set to be formed for multiple beam angles. These data can be understood as the projections of the object index, i.e. the Radon transform of the index of the object. Using filtered backprojection algorithms sets of projection data were used to reconstruct PICT images. The resulting images are free from the spatial distortions found in standard OCT. Experimental results show that PICT images correspond well with the dimensional characteristics of specific samples.