The Guayule (Parthenium argentatum, Asteraceae), is a small ramified tree native to the northern Mexico and southwestern United States. The guayule shows a growing interest in research and agriculture (Ray, 1993) due to its hypoallergenic latex properties (Taurines et al., 2019), and seems adapted to South France climate (Sfeir and al., 2014). However, the production itineraries in relation to latex production are still not assessed, and so far little studies were done on the plant structure and functioning. This study aims to propose a first FSPM of the species using the GreenLab model, calibrated from data issued from two varieties in different environmental conditions. The studying methodology is first based on a qualitative architectural analysis (Barthelemy and al., 2007). Second, on the various axis typologies, the development and branching stochastic rules can then be retrieved from field internode distributions collections. Finally, the organ source and sink relations parameters can be fitted from dedicated dry weight measurements (Kang et al., 2018). Experimental plots were hold south of France, close to Montpellier on two varieties CL1 and CLA1, with six environmental conditions related to density (9091 and 62500 plants per hectare) and hydric pressure (no stress, low stress and high stress). 50 plants per environmental conditions were measured. The sampling was optimized to the plant structure and to quantify the polyisoprene and resins contents. The guayule shows a sympodial development is composed of modules with terminal inflorescence. Its architecture corresponds to the Leeuwenberg's model (Hallé et al., 1978). The axes are constituted of successive modules. Over a year, the plant produces eight to nine successive modules. Studying the plant structure, we found out that the number of relay axis per module follows a binomial distribution. The modules are ordered from the plant base to the top. And these modules are composed of internodes whose number a