Methylene blue dye use during mid-trimester amniocentesis in twin pregnancy is associated with a high risk of small intestinal atresia. It is plausible that the effects of methylene blue as a fetotoxic agent may also lead to fetal death. We conducted a retrospective cohort study of all women who had an amniocentesis during twin pregnancy from 1980 through 1991 in New South Wales, Australia. Women who were exposed to methylene blue dye during the procedure were compared with women who had amniocentesis without dye exposure. Fetal death occurred in 31.8 per cent of pregnancies that had exposure to a high concentration of methylene blue, compared with 14.5 per cent of pregnancies exposed to a low concentration and 4.3 per cent of pregnancies with no exposure to dye. The unadjusted and adjusted risks and 95 per cent confidence intervals (CIs) for fetal death after any exposure to dye were 5.03 (2.12-11.91) and 8.52 (2.28-31.80), respectively. The adjusted odds ratio and 95 per cent CIs for the low and high concentration dye solutions were 4.63 (0.93-23.13) and 14.98 (3.40-66.08), respectively (chi-squared test for trend P < 0.001). Fetus papyraceous was significantly more likely among pregnancies exposed to a high concentration of methylene blue (P < 0.001) than among unexposed pregnancies. These results support the hypothesis that methylene blue dye use during mid-trimester amniocentesis in twin pregnancy increases the risk of fetal death.