A key uncertainty of wastewater-based epidemi-ology is the size of the population which contributed to a givenwastewater sample. We previously developed and validated aBayesian inference model to estimate population size based on14 population markers which: (1) are easily measured and (2)have mass loads which correlate with population size. However,the potential uncertainty of the model prediction due toin-sewer degradation of these markers was not evaluated. Inthis study, we addressed this gap by testing their stability undersewer conditions and assessed whether degradation impacts themodel estimates. Five markers, which formed the core of ourmodel, were stable in the sewers while the others were not.Our evaluation showed that the presence of unstablepopulation markers in the model did not decrease the precision of the population estimates providing that stable markerssuch as acesulfame remained in the model. However, to achieve the minimum uncertainty in population estimates, we proposethat the core markers to be included in population models for other sites should meet two additional criteria: (3) negligibledegradation in wastewater to ensure the stability of chemicals during collection; and (4) < 10% in-sewer degradation could occurduring the mean residence time of the sewer network.