PurposeThe mismatch of the thermal expansion coefficients of the materials in multiplayer structure may induce serious stress concentrations in electronic packaging. Experimental evaluation of the thermal stresses and strains in those electronic composites is becoming significantly important for optimizing design and failure prediction of the electronic devices.Design/methodology/approachDigital image correlation (DIC) technique was utilized to obtain thermal deformation filed of a BGA package. With the help of white light to illuminate the cross section of the BGA package, the gray images were taken from the rough surface of the specimen, that offer a kid of carrier pattern for the DIC processing with statistical resemblance in gray distributions. By using the algorithm of correlation computation, the DIC searched the matching spots in a pair of those images in which the spot displacements were involved in between, to obtain the deformation fields of the package specimen caused by temperature changes.FindingsThe results show interesting strain distributions in the assembly. Both the horizontal displacement component and its normal derivative are strongly related to the arrangement of the solder joints in the bonding medium between the die and the ceramic substrate. The strain components in the middle region of the package are larger than those in the side regions where the strain relaxation may exist near the stress‐free boundaries. The shear strain components show special bands of parallel lines with identical amount over the chip‐package to sustain the shearing of the packed structure under thermal loading.Originality/valueThe DIC technique shows to be a useful tool for the thermal strain analysis of the electronic packaging devices. Not only provides it the whole field deformation of the assembly, but also maintains the surface pictures of the package without covering any fringes, which is important to compare the deformation field with the specimen surface to reveal the stain distribution related to the failure prediction of the materials.