Current development processes for automotive Electronic Control System (ECS) architectures have certain limitations in evaluating and comparing different architecture design alternatives. The limitations entail the lack of systematic and quantitative exploration and evaluation approaches that enable objective comparison of architectures in the early phases of the design cycle. In addition, architecture design is a multi-stage process, and entails several stakeholders who typically use their own metrics to evaluate different architecture design alternatives. Hence, there is no comprehensive view of which metrics should be used, and how they should be defined. Finally, there are often conflicting forces pulling the architecture design toward short-term objectives such as immediate cost savings versus more flexible, scalable or reliable solutions. In this paper, we propose the usage of a set of metrics for comparing ECS architecture alternatives. We believe the set of metrics constitutes a relevant aspect to address the existing design gaps. We define the set of metrics based on non-functional requirements (reliability, vehicle availability, safety, monetary cost, and timing), the degree to accommodate changes (reusability, flexibility, scalability, and expandability), the customer requirements (integrity, maintainability, energy efficiency, and security), and the compatibility to legacy designs (complexity, organizational alignment, backwards compatibility, and packagability). The key pillars for the metric-based evaluation framework of ECS architecture alternatives are as follows: (a) incremental design knowledge (by a combination of prediction, measurement, and analysis of the architecture alternatives), (b) metric evaluations (ranging from qualitative to quantitative depending on the stage of the design), and (c) comparison of the alternatives (once the metrics have been evaluated). The proposed metric-based evaluation of architecture design alternatives is relative (i.e., the results are usable within the scope of a group of alternatives being considered). The comparison is potentially more qualitative in the early phases of the design cycle as fewer data are available; for a more quantitative evaluation, data sets and design details (which are usually available in the later phases of the design cycle) are essential.