It is a fairly common practice to issue biochemical results of multichannel analysers containing many analytes, irrespective of the number of tests requested by the physician. Abnormal results in such profiles may occur from presentation of artefacted blood samples; it is sometimes not possible to differentiate the abnormality arising as a result of pathological process from that due to a poor sample quality. In this study normal volunteers were studied to examine changes occurring when centrifugation was delayed for a few hours to a few days. Analyses on stored serum samples over the same period were also studied. If haemolysis is present, lactate dehydrogenase (hydroxybutyrate dehydrogenase), aspartate transaminase, potassium, chloride, bicarbonate, phosphate and creatinine estimations are not valid. Accurate data, however, can be obtained on sodium, urea, proteins, albumin, alkaline phosphatase and gammaglutamyl transferase. The very concept of efficient and cost-effective testing by multichannel profiles may generate further unnecessary investigations with a consequent waste of health service resources if pre-analytical factors are not taken into consideration.