Wastewater analysis has the potential to provide objective information on community drug use. Introducing a population biomarker (PB) in the sample analysis may significantly reduce errors in the back-calculation associated with population estimation and wastewater volume measurement. A number of potential PBs have been suggested but no systematic evaluation has been conducted so far. This study evaluated the eligibility of the previously suggested PB candidates (creatinine, cholesterol, coprostanol and cotinine) as well as three new ones (cortisol, androstenedione and 5-hydroxyindoleacetic acid (5-HIAA)) using five criteria. We assessed the quantification method, affinity to particulate matter and stability of candidates in wastewater, as well as the constancy of inter-day excretion and correlation between excretion and census population. All PB candidates were quantifiable in wastewater. Cholesterol and coprostanol were eliminated from further consideration due to affinity to particulate matters in the wastewater. Creatinine, cortisol and androstenedione were disqualified for stability reasons. On a population scale, both cotinine and 5-HIAA were excreted (RSD=8.01±1.13% and 10.20±0.89%, respectively) at a constant rate and concentrations of each correlated well with the census population (r=0.9809 and 0.9442, respectively). Overall, both cotinine and 5-HIAA are eligible PBs, but the neurotransmitter metabolite 5-HIAA may be more suitable for international comparisons.