Mechanical and thermal characterization of polymeric materials is critical to "predictive modeling" required for reliability assessment and design optimization at the conceptual stage of a new product development. One of the most critical polymeric properties required for predictive stress modeling is "chemical shrinkage" developed during the polymerization process. Residual stresses induced by the chemical shrinkage can largely influence the performance and reliability of packages [1].The issue of "chemical shrinkage" is not new. Numerous testing methods have been developed for many decades and some of them are practiced routinely to measure the intrinsic (or total) chemical shrinkage of polymers. It is important to note, however, that not all of the intrinsic chemical shrinkage contributes to the residual stresses simply because some of chemical shrinkage occurs before the gelation point where polymers start to build mechanical strength [2].This study proposes an innovative measurement method to measure the effective chemical shrinkage. The method is based on a fiber Bragg grating (FBG) sensor. A polymer of interest is cured around a glass FBG and the Bragg wavelength shift is continuously measured and documented while polymerization progresses at the curing temperature. This configuration is referred to as Configuration I [3]. From the theoretical relationship between the Bragg wavelength shift and the effective chemical shrinkage, the evolution of the effective chemical shrinkage is determined by the measured Bragg wavelength shift.Another important property that is required for accurate determination of the residual stresses is Young's modulus, which changes non-linearly during the shrinkage evolution. Without the information about the modulus evolution, the residual stress produced by the chemical shrinkage cannot be determined accurately. A physical configuration different from one that was used for chemical shrinkage is utilized to determine the modulus evolution (this configuration is referred to as Configuration II). In this configuration, Bragg wavelength changes with both effective chemical shrinkage and modulus. The modulus evolution is then determined from the measured Bragg wavelength shift using the effective chemical shrinkage that has been determined previously from Configuration I.The proposed method was implemented for an underfill material. In the configuration I, the radius of the polymer cured around the FBG was 200 times larger than the radius of the fiber, i.e., 12.5 mm. The experimentally measured Bragg wavelength evolution during the curing process is shown in Fig.1. The effective chemical shrinkage determined from the data is shown in Fig. 2. The measured shrinkage evolution was fitted nicely by the general kinetic equation for epoxies [4]. Bragg Wavelength (nm) Time (min) 1549.112nm ( 175 o C ) Fig.1 Bragg wavelength shift of Configuration I during polymerization. 0 4 0 8 0 1 2 0 1 6 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Chemical shrinkage (%) Time (min) Measured Predicted Fig.2 Evolution of ...