The Last Planner System (LPS) constitutes a systematic method for planning and control based on the generation of short-term commitments by the workforce and the weekly control of their accomplishments in search of continual improvement. This approach allows for the stabilization of workflow and uncertainty reduction in short-term plans, which are assessed using the Plan Percent Complete (PPC) indicator and the systematical collection of Reasons for Non-Compliance (RNC). Our research goal is to contribute to the understanding of how PPC and RNC metrics can be used for early assessment of project performance concerning schedule accomplishment. We used a sample of 25 Chilean projects with weekly information regarding progress, PPC, RNC and time deviation, that was categorized into two groups according to their schedule accomplishment results, using a clustering algorithm. We compared the PPC and RNC indicators from the two groups across project execution to detect significant differences. We found that successful projects evidence a statistically significant increase in the PPC, compared to the less-than successful group, lower PPC variability and a lower number of RNC per week, since early project execution. The results allowed us to conclude that these metrics can help perform early assessments of project performance.