The migration of low molecular weight organic compounds through polyethylene terephthalate (PET) films was determined by using a custom permeation cell assembly. Fatty food simulant (Miglyol 812) was added to the receptor chamber, while the donor chamber was filled with 1% and 10% (v/v) migrant compounds spiked in simulant. The permeation cell was maintained at 40°C, 66°C, 100°C or 121°C for up to 25 days of polymer film exposure time. Migrants in Miglyol were directly quantified without a liquid-liquid extraction step by headspace-GC-MS analysis. Experimental diffusion coefficients (DP) of toluene, benzyl alcohol, ethyl butyrate and methyl salicylate through PET film were determined. Results from Limm's diffusion model showed that the predicted DP values for PET were all greater than the experimental values. DP values predicted by Piringer's diffusion model were also greater than those determined experimentally at 66°C, 100°C and 121°C. However, Piringer's model led to the underestimation of benzyl alcohol (Áp = 3.7) and methyl salicylate (Áp = 4.0) diffusion at 40°C with its revised "upper-bound" Áp value of 3.1 at temperatures below the glass transition temperature (Tg) of PET (<70°C). This implies that input parameters of Piringer's model may need to be revised to ensure a margin of safety for consumers. On the other hand, at temperatures greater than the Tg, both models appear too conservative and unrealistic. The highest estimated Áp value from Piringer's model was 2.6 for methyl salicylate, which was much lower than the "upper-bound" Áp value of 6.4 for PET. Therefore, it may be necessary further to refine "upper-bound" Áp values for PET such that Piringer's model does not significantly underestimate or overestimate the migration of organic compounds dependent upon the temperature condition of the food contact material.