Maternal stress is commonly cited as an important risk factor for spontaneous abortion. For humans, however, there is little physiological evidence linking miscarriage to stress. This lack of evidence may be attributable to a paucity of research on maternal stress during the earliest gestational stages. Most human studies have focused on "clinical" pregnancy (>6 weeks after the last menstrual period). The majority of miscarriages, however, occur earlier, within the first 3 weeks after conception (approximately 5 weeks after the last menstrual period). Studies focused on clinical pregnancy thus miss the most critical period for pregnancy continuance. We examined the association between miscarriage and levels of maternal urinary cortisol during the first 3 weeks after conception. Pregnancies characterized by increased maternal cortisol during this period (within participant analyses) were more likely to result in spontaneous abortion (P < 0.05). This evidence links increased levels in this stress marker with a higher risk of early pregnancy loss in humans.