Our long-range goal is to improve current whole-heart CT calcium score by extracting quantitative features from individual calcifications. We performed deconvolution to improve small calcifications assessment which challenge conventional CT calcium score scanning resolution. We analyzed features of individual calcifications on repeated standard (2.5-mm) and thin (1.25-mm) slice scans from QRM-Cardio phantom, cadaver hearts, and CARDIA study participants. Pre-processing to improve resolution involved of Lucy-Richardson deconvolution with a measured PSF or 3D blind deconvolution where the PSF was iteratively optimized on high detail structures like calcifications in the images. Using QRM with inserts having known mg-calcium, we determined that both blind and conventional deconvolution improved mass measurements nearly equally well on standard images. Further, de-convolved thin images gave excellent recovery of actual mass scores, suggesting that such processing could be our gold standard. For CARDIA images, blind deconvolution greatly improved results on standard slices. Accuracy across 33 calcifications (without, with deconvolution) was (23%,9%), (18%,1%), and (-19%,-1%), for Agatston, volume, and mass scores, respectively. Reproducibility was (0.13,0.10), (0.12,0.08), and (0.11,0.06), respectively. Mass scores were more reproducible than Agatston scores or vol-ume scores. Cadaver volumes showed similar improvements in accuracy/reproducibility and slightly better results with a measured PSF. For many other calcification features in CARDIA data, blind deconvolution improved reproducibility in 21 out of 24 features. Deconvolution improves accuracy and reproducibility of multiple features extracted from individual calcifications in CT calcium score exam. Blind deconvolution improves feature assessments of coronary calcification in archived datasets.