In 1999, the Canadian Perinatal Surveillance System of Health Canada decided to undertake a national survey of Canadian women's experiences of their pregnancy, birth and postpartum care. The challenges encountered in selecting a representative sampling frame and developing a sound methodology for conducting a survey of Canadian women at six months after birth are addressed. We considered the advantages and disadvantages of six different sampling options. A sample based on the Census emerged as the optimal approach for providing the most reliable and representative sample.