A useful analytical method of measuring reaction kinetics of polymerization is differential scanning calorimetry (DSC). One problem in DSC testing of free radical polymerization kinetics, however, is that the induction time before polymerization begins varies inversely with sample size. This paper shows that this "sample size anomaly" results from a synergistic reaction between oxygen and certain inhibitors. Two types of inhibitors were found: oxygen-inert inhibitors such as benzoquinone which do not react with oxygen and hence show little sample size anomaly, and oxygen-active inhibitors such as hydroquinone which require oxygen to be active. A model describing the effect of oxygen during inhibition was verified using isothermal DSC data for styrene homopolymerization. The inhibition time was also measured for a variety of sample pans using a commercial vinyl ester resin formulation, and observed behavior indicates the presence of both oxygen-active and oxygen-inert inhibitors. The recommended procedure for DSC sample preparation is to match the polymerization conditions during processing. In composites processing, for example, resin is saturated in air and cured in a closed mold environment, so the appropriate DSC sample procedure is to completely fill the sample pan to displace air in the headspace without changing oxygen concentration in the sample.