Using the population-based Connecticut Tumor Registry, postmastectomy breast reconstruction was examined in the 10,756 breast cancers in Connecticut women (including 10,133 in white women and 554 in black women) diagnosed from 1988 to 1995. Reconstruction increased from 6.4 percent of cancers in 1988 to 9.1 percent in 1991, but it declined to 4.7 percent of cancers in 1992 (when the Food and Drug Administration instituted a restriction on the use of silicone gel implants); by 1995, the rate had recovered to 8.5 percent. Reconstruction was negatively associated with age, poverty rate of the census tract of residence, and black (versus white) race; these associations require further study.