Nearly all field experiments on food crops which are conducted in the Caribbean region employ techniques of design and layout which have been developed by biometricians in temperate countries in North America and Europe. These techniques have been very successful for field experimentation in the countries in which they were developed. However they have not been so successful in tropical regions_ This is partly due to the fact that often the land used for experimentation in the tropics is not uniform which means that great care has to be taken to ensure that each block is laid down so as to eliminate most of the variations due to non uniformity of land_ Again the type of factorial arrangements often chosen include treatment combinations which are not really of interest to the experimenter. Researchers are therefore urged to adopt a flexible approach to the design of their experiments so as to include only treatment combinations in which they are interested and also to give extra replication to treatments which are of special interest such as control treatments. Another problem regularly faced by experimenters is to decide on the size and shape of their plots. Some suggestions are made to help in choosing the optimum plot size and the advantages of various plot shapes are also discussed.