OBJECTIVES: We aimed to calculate analytical variation and reference change value for biochemistry analytes studied on the Roche Cobas 8000 analyzer that we routinely use in our laboratory.MATERIALS and METHODS: The reference change value for our analytes;calculated with formula 2 1/2 x Z x (CV A 2 + CV I 2 ) 1/2 .( 95% probability, two-way; Z=2.58 taken) Three-month, two-level internal quality control results were used to calculate analytical coefficient of variation (CV A ). Median values from meta-analysis studies in EFLM biological variation database were used for intraindividual (CV I ) and inter-individual (CV G ) coefficients of variation.The formulas CVA < 0.25 x CVI for optimum performance, CVA < 0.50 x CVI for acceptable performance and CVA < 0.75 x CVI for minimum performance were used to determine quality specifications developed for precision.RESULTS: When evaluated accordingly, the reference change values (%) for our analytes are in order; glucose 16, urea 39, creatinine 19, total cholesterol 15, triglyceride 55, HDL-cholesterol 19, LDL-cholesterol 24, AST 27, ALT 29, sodium 5, potassium 13, chlorine 8, calcium 7, magnesium 10, phosphorus 23, iron 57 calculated.Optimum for urea, triglyceride, AST and iron from our analytes; acceptable for total cholesterol, LDL-cholesterol, ALT and phosphorus; it meets the minimum performance quality specifications f or g lucose, HDL c holesterol a nd potassium; creatinine, sodium, chlorine, calcium and magnesium do not seem to provide.CONCLUSIONS: Analytical and biological sources of variation should be considered for reliability of laboratory results.Tests that do not meet quality specifications require root cause analysis followed by corrective and preventive action.