A classical sintering theory predicts that a pore with high coordination number grows instead of shrinking during sintering. This has led to the proposition that grain growth may be beneficial to densification. Pan and colleagues argue against this theory using computer simulations, while Flinn and colleagues have provided direct experimental evidence showing that very large pores shrink despite their high coordination number. The current paper brings the analytical and experimental work together. In particular, further computer simulation evidence is provided to support the argument and an analytical model is developed to predict the densification rate of powder compacts containing large and small pores. The analytical model shows that there is a sudden reduction in the shrinking rate of the large pores when the small pores are eliminated.