In arid regions characterized by large variations in groundwater salinity, the data derived from irrigation and domestic water supply wells may exhibit bias, reflecting an overall lower salinity than the true aquifer distribution. This bias stems from the decommissioning, non-use, or disrepair of wells that are frequently sources of higher salinity readings, rendering them unavailable for sampling. Baseflow-fed streams, agricultural drains, seeps, springs issuing into agricultural drains, and randomly located test hole samples tend to manifest higher averages and ranges of salinity when compared to supply wells. Agricultural drain flows, springs, and test holes, if sampled following recommended guidelines, are less susceptible to such bias. This study presents a case of groundwater bias identified through an initial water well sampling program in El Paso (Texas, USA). Subsequent rounds of sampling, incorporating drain samples, spring samples, and test hole samples, revealed a more comprehensive understanding of the salinity dynamics. The dataset not only highlights the existence of bias but also provides evidence for a combined geological and agricultural origin of salinity. Additionally, it demonstrates that drain sampling in an earlier study did not accurately depict a primary salinity source due to incomplete analysis of the data. Recommendations are outlined to mitigate bias, emphasizing the importance of sample control from baseflow-fed drains, springs, water wells, and test hole samples. The study also infers the upwelling of saline groundwater from deeper formations in the study area, contributing to a more comprehensive understanding of groundwater salinity dynamics.