To better understand 'within subject' variability in masticatory performance of artificial food, this study evaluates the effects of bolus size and chewing rate. Twenty young adult male subjects each participated in two experiments evaluating their ability to process an artificial food (CutterSil). Based on the frequency distributions for the weights of the chewed particles, median particle size and broadness of the particle distribution were estimated. Controlling for the number of chews, the effect of (1) bolus size was evaluated by having each subject chew one whole tablet, four quarters, three quarters or two quarters; (2). chewing rate was evaluated by having the subjects chew first habitually, then at slow (40 cycles/min) and fast (100 cycles/min) rates. The results showed that median particle sizes decreased and the distribution of particles widened with decreasing bolus size. The slow chewing rate produced the smallest particles with the widest distributions, followed by the habitual and fast rates, respectively. The study has shown that bolus size and chewing rate are important source of variation which must be considered in studies of masticatory performance.