The usual methods for site evaluation, and later the determination of effective operation of an on-site wastewater system, whether a traditional septic drainfield or a modern aerated wastewater system, are for soil and effluent samples to be taken and compared against some arbitrary ideal. Regulators seem keen for check lists of soil, water and waste constituents for analysis either at the planning phase or later to 'check' on the operational performance and sustainability of the system. Whether these activities have any real role in risk assessment and environmental security, the variability between and within samples needs to be considered prior to preparation of monitoring routines. This paper examines potential variability in soil sampling and analysis depending upon the location of the sample, the preservation of the sample, deterioration of samples and the analytical results. While it is recognised that in-soil variability is high, there are erroneous expectations that sample analyses are precise and accurate. Within and between sample results and within and between laboratory results often fall outside an expected range of variability. This paper uses data from national and international inter-laboratory proficiency testing programs in soil and water analysis to examine typical variability for identical samples. The consequences of assuming high levels of accuracy may be misconceived and the end user needs to be conscious of possible natural variations. When the variability of sampling is added to this assessment, wide ranges in typical values of contaminants, pollutants and nutrients can be expected.