Mining of genetic diversity for rice molecular breeding, through discovery and utilization of useful genes and alleles, holds promise for ensuring global food security for systems challenged by climate and environmental change. Genome-wide association studies (GWAS) are a first and necessary step to (1) identify candidate loci/genes associated with desirable traits, and (2) assess allelic diversity on such loci and identify putative donor materials carrying it. GWAS requires phenotyping of large populations. The GRiSP Global Phenotyping Network implements this approach on a multi-partner basis well beyond CGIAR, enabling multi-phenotyping of agreed-on panels established for major genetic groups of rice. Multi-phenotyping in many environments and for many traits provides important add-on information, e.g. on G x E interactions (for a given trait) and cross adaptations (co-localization of different traits). The network also combines field-based and controlled-environment approaches, as well as crop-model-assisted (heuristic) approaches for 'hidden' physiological traits. Initiated in 2012, it is now building a large phenomics database for traits related to yield potential and abiotic stresses (in the future possibly also biotic stresses and grain quality traits). This database will also include phenomics resources previously generated by CIRAD's ORYTAGE (Diversité des caractères d'adaptation aux contraintes hydriques et thermiques chez le riz) project and the Generation Challenge Program (GCP). GWAS will start in late 2013 when the panels have been genotyped and a critical mass of phenomics data is available, and will continue as a rolling process as the database grows. The results will be shared by the entire network, while giving individual owners the opportunity to publish them first, and will then be in the public domain. This paper highlights early phenomics results on indica-type rices from the PRAY (Phenomics of Rice Adaptation and Yield) and ORYTAGE projects, w