A Cochrane review included 2 trials involving 1689 women. The contraceptive vaginal sponge made of polyurethane impregnated with spermicide nonoxynol-9 was significantly less effective in both trials in preventing overall pregnancy than was the diaphragm (used with nonoxynol-9). In the larger US trial, the 12-month cumulative life-table termination rates per 100 women for overall pregnancy were 17.4 for the sponge and 12.8 for the diaphragm. The rates were 24.5 for the sponge and 10.9 for the diaphragm in the UK trial. Similarly, discontinuation rates at 12 months were higher with the sponge than with the diaphragm (OR 1.3; 95% CI 1.1–1.6). Allergic-type reactions were more common with the sponge, although discontinuation for discomfort differed in the two trials.