Urine is recognized as the prime matrix for drug test screening with well-established methods and testing protocols. Its major limitation is with regard to the inconvenience of sample collection and lack of integrity due to adulteration, dilution, drug spiking or sample exchange. The question is whether oral fluid, with its apparent better sample integrity, can replace urine for drug screening. This review examines the sample integrity problems and the advantages and limitations of oral fluid and urine in drug screening programmes. The variety of sample collection devices for oral fluid is shown to be a problem with recovery and detection for some drugs. This is examined in relation to the pharmacokinetics of drug metabolism and excretion in this matrix. Buccal contamination with drugs in oral fluid may also cause problems with interpretation. The clinical advantages of oral fluid analysis compared with urine testing are highlighted. Parent drugs are often found in oral fluid where only their metabolites may be found in urine, for example the benzodiazepines. 6-Monoacetylmorphine, an indicative marker of heroin, has a high prevalence in oral fluid from users of this drug but its detection in urine is limited due to its short half-life. Advances in analytical techniques, particularly chromatography linked to tandem mass spectrometry, are helping to promote oral fluid analysis. However, the lack of concordance studies examining both urine and oral fluid drug levels and kinetics in the clinical setting is of some concern.