The aim of this study was to investigate whether a strong and clinically applicable correlation exists between saliva and whole-blood tacrolimus levels measured by high-performance liquid chromatography-tandem mass spectrometry. A high degree of correlation would potentially allow pain-free saliva sample collection to replace blood sampling for the measurement of tacrolimus levels. Enrolled in the study were 37 children (24 boys) aged 8-18 years [median (IQR) 16.2 (12.9-17.5) years] attending the renal transplant clinic at the Royal Manchester Children's Hospital and 77 paired blood saliva samples were collected. The mean (SD) saliva tacrolimus level was 0.14 (0.16), range 0-0.7 μg/l. In ten cases, tacrolimus was not detected in the saliva despite being present in blood. The ratio of blood-to-saliva tacrolimus levels varied from 2.6 to 550. The Pearson product-moment correlation suggested a weak linear relationship between tacrolimus levels in blood and saliva with a coefficient 0.36. Individual patients did not demonstrate consistent tacrolimus blood/saliva ratios with a mean correlation of 0.08. Additional experiments excluded saliva contamination with blood and sample collection and storage conditions as causes of poor correlation. The measurement of saliva tacrolimus levels in place of or as an adjunct to blood sampling therefore cannot be recommended.