Insulin resistance is implicated in the pathogenesis of preeclampsia, but prospective data are limited. SHBG, a marker of insulin resistance among nonpregnant individuals, has not been studied in detail during pregnancy. We conducted a prospective, nested, case-control study to test the hypothesis that increased insulin resistance, marked by reduced first trimester SHBG levels, is associated with increased risk of subsequent preeclampsia. First trimester SHBG levels were measured in 45 nulliparous women who subsequently developed preeclampsia (blood pressure, > or =140/90 mm Hg; proteinuria, either > or =2+ by dipstick or > or =300 mg/24 h, after 20 wk gestation) and in 90 randomly selected normotensive nulliparous controls. Compared with controls, women who developed preeclampsia had significantly reduced first trimester SHBG levels (302 +/- 130 vs. 396 +/- 186 nmol/liter; P < 0.01). Every 100 nmol/liter increase in SHBG was associated with a 31% reduced risk of preeclampsia [odds ratio (OR), 0.69; 95% confidence interval (CI), 0.55, 0.88; P < 0.01]. After adjusting for covariates in a multiple logistic regression model, the association between first trimester SHBG and preeclampsia remained significant (per 100 nmol/liter increase; OR, 0.66; 95% CI, 0.47, 0.92; P = 0.01). When subjects were stratified by body mass index (lean: body mass index, < 25 kg/m(2); overweight: body mass index, > or =25 kg/m(2)), overweight women had lower SHBG levels than lean women (286 +/- 156 vs. 410 +/- 166 nmol/liter; P < 0.01), and within each stratum, women with preeclampsia had lower SHBG levels than their respective controls. In a multivariable analysis, the association between SHBG and preeclampsia strengthened among lean women, such that every 100 nmol/liter increase in serum SHBG was associated with a 55% reduction in the risk of preeclampsia (OR, 0.45; 95% CI, 0.27, 0.77; P < 0.01), whereas in overweight women, the association was mitigated (OR, 1.02; 95% CI, 0.62, 1.69; P = 0.9). We conclude that increased early pregnancy insulin resistance is independently associated with subsequent preeclampsia. First trimester SHBG levels may be a useful biomarker for preeclampsia, especially among lean women who otherwise would be perceived to be at low risk.