Few layers boron nitride nanosheets (BNNSs) were derived from precursors of specially synthesized micro-sized hexagonal boron nitrides (h-BNs) of large flakes through microfluidization. They were then considered promising filler constituent for improving creep and recovery behaviors of low density polyethylene (LDPE) films. To enhance the compatibility between BNNSs and LDPE matrix, attachment of trimethoxy silane (VTS) coupling agent to the surface of BNNSs was performed through microfluidization, followed by chemically induced graft copolymerization of 2-hydroxyethyl methacrylate (HEMA) onto the surface of LDPE beads by utilizing benzoyl peroxide as initiator. LDPE-g-HEMA films containing various amounts (0.3, 0.5 and 1 wt.%) of silanized BNNSs were consecutively produced, conducting melt extrusion. Neat LDPE films and LDPE-g-HEMA films with and without untreated BNNSs (1 wt.%), were also produced in the same way for the sake of comparison. Transmission electron microscopy (TEM) examination showed that morphology of BNNSs was altered after silane treatment, and that silanized BNNSs were dispersed in the form of exfoliated sheets within the grafted film. Creep and recovery behaviors of the resulting films were determined using a Thermal Mechanical Analyzer (TMA). As a result, regardless of the weight content, the grafted films with silanized BNNSs were found to exhibit low creep recovery compliance combined with low maximum creep strain as compared to neat LDPE films and the grafted films with untreated BNNSs, suggesting that mobility of the grafted matrix molecules has become more restricted when they were brought in contact with the silane molecules on the surface of BNNSs. Various different models, including the Findley's power law, Burgers, and Weibull distribution were then proposed to fit the experimentally determined creep and recovery data in order to gain more insight into the effect of untreated and silanized BNNSs on the film deformation mechanisms. The findings obtained were then briefly discussed with respect to theoretical predictions.