Studies into the risks of environmental hazards on human health study require accurate estimates of exposures that might be experienced by the populations at risk, and this information often comes from monitoring networks. However, there may be periods of missing data and in many epidemiological studies the locations and times of exposure measurements and health data do not match, due to them coming from completely different sources. Further, monitoring sites in environmental networks may be preferentially located by the designers, e.g. to monitor potential “hot spots”, meaning that the data arising from such networks may not accurately characterize the spatio–temporal field they intend to monitor, leading to biased estimates of exposures, and subsequently to biased estimates of risks. Here, we will illustrate the potential effects of preferential sampling using examples from air pollution and health, and discuss possible strategies for mitigating those effects in this setting.