The architectural mathematical model Greenlab was tested for assisting plant phenotyping on two rice genotypes: Nippon Bare (NB) and Phyllo, its organogenesis mutant. A 50 days experiment was realized to compare, at organ level on NB, tiller pruned NB and Phyllo, morphogenesis, sugar partitioning and assimilation rate. Phenotypic data were used to optimize Greenlab parameters, considered then as process based traits. Experimental analysis showed weak modifications of pruned rice morphogenesis occurred. In contrast, Phyllo showed quicker organogenesis, small organs, no tillering, half the assimilation rate of NB, modified root sugar metabolism. Greenlab fittings on observations were good except for NB, mainly because of the difficulty of correctly optimizing numerous small leaf sinks (generated by process) compared to a single big root sink compartment. However, Greenlab successfully sensed the phenotypic differences underlined experimentally, in particular, modified carbon sink to source relationships for Phyllo. Modeling analysis of rice tillering cases shows that Greenlab has a great potential as heuristic approaches to be applied to plant phenotyping, more case studies are proposed to testing the model as a suitable tool to assist to screen plant phenotype.