The standard formulas used to calculate sample size for an individually matched case-control study assume a constant probability of exposure throughout the pool of possible controls. We propose new ...