The evaluation was able to draw upon a population of automobiles that never received either parts marking or anti-theft systems, which acted as a comparison group. The analysis used statistical procedures, based on Generalized Linear Least Squares and Maximum Likelihood Poisson regression models, to estimate the reduction in automobile theft attributable to parts marking and factory-installed anti-theft devices. The application of cost-benefit criteria required that the estimated reduction in theft be valued at the cost of a stolen car compared with the cost of marking automobile parts and equipping them with anti-theft equipment at the factory. The evaluation concluded that both parts marking and anti-theft devices are cost-effective ways of reducing automobile thefts. One estimate concluded that parts marking reduces automobile theft by 138 to 207 cars per 100,000 registered automobiles. Another estimate indicated that parts marking reduces car theft by 220 to 300 automobiles per 100,000 registered cars. Given the benefit from preventing an auto theft ($6,000 on average) and assuming the cost of marking automobiles at $0.50 per car per year, parts-marking would be cost-beneficial if it prevented 9 thefts per 100,000 registered automobiles. Regarding the effectiveness of anti-theft devices, one estimate indicates that these devices reduce auto thefts by between 311 and 445 per 100,000 registered cars per year. Another estimate is theft reduction by between 413 and 475 per 100,000 registered cars. Assuming a $6,000 average loss per stolen car and assuming that anti-theft devices cost $20 per car per year, anti-theft devices would have a favorable benefit-cost ratio if the devices prevented 330 thefts per 100,000 registered cars per year. Extensive tables and figures and 24 references