Purpose: Study the trade‐offs between lesion detectability and dose to radiosensitive organs during angiographic procedures using Monte Carlo techniques for quantitatively estimating peak/average organdoses and an implementation of the Hotelling observer. Methods: We performed a signal‐known‐exactly/background‐known‐exactly experiment by using a high‐resolution cardiac phantom with detailed coronary artery models (freely available at hades.googlecode.com) registered in a human body phantom. A set of stenoses with varying sizes were synthesized and placed on the coronaries. To consider the scatter effect, we applied the Monte Carlo method (freely available at penmesh.googlecode.com) to simulate angiographic procedure for different X‐ray tube currents and peak voltages with an ideal detector model. Projections, pixel uncertainty estimates and peak/average organdoses were determined for each simulation setting. We repeated the simulation for 50 times with the exact particle number and the same scenario. By this mean, the projection pixel values for the same detector bin are under Poisson distribution. Finally, we calculated signal‐ noise‐ratio (SNR) (to be available at mumoc.googlecode.com) by using a Hotelling observer from projection images to assess signal detectability.Results: We estimated 11 peak organdoses normalized by SNR2 for a 6.86‐mm3 stenosis at 60/90/120 kVp. Our findings suggest that the spinal cord is the most exposed organ with peak doses increasing by 4.6 times when the tube voltage increases from 60 to 120 kVp. For a constant SNR, peak/average organdoses increase 1.1/1.4 times on average between 60 and 90 kVp, and 3.0/4.2 times on average from 60 to 120 kVp. Conclusions: We investigated peak organdoses for deterministic radiation injuries in cardiac angiographic procedures. We found that lower X‐ray tube voltages (60 to 90 kVp) produce higher SNR than 120 kVp. The methodology presented here can be applied to the optimization of X‐ray imaging systems for other radiological procedures besides angiography.