Plating techniques for microbial enumeration continue to be important quantitative performance assessment
tools used during drinking water compliance testing (e.g. routine coliform analysis)
and bench- and pilot-scale performance demonstrations (e.g. aerobic spore removal). While
considerable statistical analysis of these methods historically has been limited to quality control
and method development, the analysis of experimental data and laboratory proficiency can also
be enhanced by the correct analysis of variability in the data. This paper presents a simple
framework for assessing method-associated errors in the quantification of bacteria by membrane
filtration - agar plate methods and provides a preliminary discussion of probabilistic models that
may be used to quantify uncertainty in plate-count data. Simple recovery experiments for
Bacillus subtilis spores and Escherichia coli are presented to demonstrate the occurrence and
quantification of method-associated errors. This type of analysis of method errors is essential for
laboratories, water suppliers, and engineers, particularly in demonstrating regulatory compliance. Includes 11 references, tables.