Benchmarking the dynamic performance of water companies is a powerful tool to develop sound policies. However, it requires the use of reliable and robust methods and quality of data. In this study, the stochastic nonparametric envelopment of data (StoNED) method was employed to assess productivity change and eco-productivity change of a sample of Chilean water companies over the period 2010–2018. The StoNED approach overcomes the main limitations of stochastic frontier analysis and data envelopment analysis techniques which are the most commonly used to evaluate performance of water companies. Productivity change metric focused on the economic performance, whereas eco-productivity change evaluated economic and quality of service performance by including non-revenue water and unplanned water supply interruptions. Hence, this study analyzed the impact of quality of service variables on the dynamic eco-efficiency of water companies. The annual average productivity change and eco-productivity change estimations were 0.18% and −0.14%, respectively. It was evidenced that on average the Chilean water companies have put as priority to reduce operational costs than to improve quality of service. In both performance metrics, efficiency change was the main contributor to performance changes, whereas the role of technical change was negligible. It was illustrated that full private water companies performed better than concessionary companies and the public company. Reducing non-revenue water and unplanned water supply interruptions was particularly costly for concessionary water companies. The methodology and conclusions of this study are of great interest for water regulators and water utilities to improve the benchmarking process in the provision of drinking water services.