To investigate whether women with unexplained recurrent pregnancy loss have impaired arterial vascular health compared with controls, and to evaluate whether this is modifiable by exercise. Experimental case-control pilot study. University medical centre in the Netherlands. Twelve women with unexplained recurrent pregnancy loss, 11 nulliparous women and 19 primiparous women with a history of uncomplicated pregnancies. In all three groups we measured baseline characteristics, metabolic components and arterial vascular health, and repeated this in women with unexplained recurrent pregnancy loss after 1 month of protocolled and supervised cycle training. Differences in arterial vascular health between women with unexplained recurrent pregnancy loss and controls, and the effect of exercise on arterial vascular health in women with unexplained recurrent pregnancy loss. Women with unexplained recurrent pregnancy loss have a significantly increased carotid intima media thickness in comparison with both controls (both P < 0.01), a significantly decreased brachial endothelial dependent flow-mediated vasodilation in comparison with both controls (nulliparous: P < 0.01; primiparous: P = 0.05) and a significantly decreased femoral endothelial dependent flow-mediated vasodilation in comparison with primiparous women (P = 0.01). The endothelium independent glyceryl trinitrate response was similar in all groups. With 1 month of exercise, the carotid intima media thickness decreased significantly by 7% (P = 0.05) and the femoral FMD increased significantly by 10% (P = 0.01) in women with unexplained recurrent pregnancy loss. Women with unexplained recurrent pregnancy loss have an impaired vascular health in comparison with controls. This impaired arterial vascular health can be improved by exercise.