Using data from the 2009 China Health and Nutrition Survey, this study investigates China's income-health gradient by analyzing the effect of both current and long-term household income on 22 blood-based biomarkers, 4 used as individual variables and all 22 assessed as a composite. After applying a two-step residual inclusion estimator, we find limited evidence of an income-health gradient irrespective of whether the income measure is current or long term. Because risky behavior may attenuate income's positive effects on health, we also analyze the association between income and such health-influencing factors as alcohol consumption, smoking, diet, physical activity, and dietary knowledge. Although we find that higher incomes go hand-in-hand with some of these factors (in particular, a higher number of cigarettes smoked per day), they also promote poorer diets (higher consumption of fats and calories). However, the fact that these effects are small, dependent on income measure, and susceptible to reporting biases makes it unlikely that they are attenuating income's potentially positive effects on health.