Signalizing an intersection usually results in a reduction in right-angle and left-turn crashes, and an increase in rear-end crashes. This study used the conventional and Empirical Bayes (EB) before and after methods on 45 treated sites (converted from stop to signal control) in California and Minnesota to estimate the safety effect of having signals installed. The results confirm the belief that right-angle and left-turn crashes are reduced and rear-end crashes increase. However, these effects cannot be used to quantitatively assess the benefit gained from the reduction in right-angle and left-turn crashes against the increase in rear-end crashes, simply because crash types have different severities. By performing an economic examination of the safety effects, this study was able to show that by installing signals on 45 treated sites, there was a positive aggregate economic benefit of $155,883,978 which represents a 69 percent reduction in cost. This translated into $616,142 per site-year.