Based on the results of sampling experiments performed on a map of sedimentary environments of the Mississippi Delta region of south-east Louisiana, it is concluded that systematic sampling is the most efficient sampling strategy overall for mapping sedimentary facies. This is true whether or not prior information about the spatial structure of the underlying pattern has been obtained. For the map area subdivided into 4034 separate sampling cells, systematic sampling based on 100 point-samples achieved an accuracy in the estimated map close to 63%. Sampling experiments which made use of spatial correlation existing in the underlying pattern achieved no higher accuracies for the same size sample. Only when virtual knowledge of the underlying pattern was assumed did the accuracy achieved increase notably, namely, 71% for sample size of 100. In future, it is recommended that sediment mapping programs be conducted wherever possible by sampling along a grid.