Understanding the excretion and treatment of human waste (feces and urine) in low and middle income countries (LMICs) is necessary to design appropriate waste management strategies. However, excretion and treatment are often difficult to quantify due to decentralization of excreta management. We address this gap by developing a mechanistic, stochastic model to characterize phosphorus, nitrogen, biochemical oxygen demand (BOD), and fecal coliform pollution from human excreta for 108 LMICs. The model estimates excretion and treatment given three scenarios: (1) use of existing sanitation systems, (2) use of World Health Organization-defined "improved sanitation", and (3) use of best available technologies. Our model estimates that more than 10(9) kg/yr each of phosphorus, nitrogen and BOD are produced. Of this, 22(19-27)%, 11(7-15)%, 17(10-23)%, and 35 (23-47)% (mean and 95% range) BOD, nitrogen, phosphorus, and fecal coliforms, respectively, are removed by existing sanitation systems. Our model estimates that upgrading to "improved sanitation" increases mean removal slightly to between 17 and 53%. Under the best available technology scenario, only approximately 60-80% of pollutants are treated. To reduce impact of nutrient and microbial pollution on human and environmental health, improvements in both access to adequate sanitation and sanitation treatment efficiency are needed.