This paper provides a model with which to predict soil ingestion recovery values in soil ingestion studies either retrospectively or prospectively. The predictive equations generated from the model can be used to estimate minimum soil ingestion detection levels from soil ingestion studies which use mass-balance methods. The model is derived from data assessing soil recovery efficiencies in adults using eight different predictive tracer elements. The results constitute a methodology for determining minimum detection levels of soil ingestion and hence have important regulatory significance.