The potential to use oral fluid as a drug-testing specimen has been the subject of considerable scientific interest. The ease with which specimens can be collected and the potential for oral fluid (OF) drug concentrations to reflect blood-drug concentrations make it a potentially valuable specimen in clinical as well as forensic settings. However, the possible effects of the OF collection process on drug detection and quantification has often been over looked. Several studies have documented that drug-contamination of the oral cavity may skew oral fluid/blood drug ratios and confound interpretation when drugs are smoked, insufflated or ingested orally. OF pH is predicted to have an effect on the concentration of drugs in OF. However, in a controlled clinical study, the effect of pH was less than that of collection technique. Mean codeine OF concentrations in specimens collected a non-stimulating control method were 3.6 times higher than those in OF collected after acidic stimulation. Mean codeine concentrations were 50% lower than control using mechanical stimulation and 77% of control using commercial collection devices. Several factors should be considered if a commercial OF collection device is used. In vitro collection experiments demonstrated that the mean collection volume varied between devices from 0.82 to 1.86 mL. The percentage of the collected volume that could be recovered from the device varied from 18% to 83%. In vitro experiments demonstrated considerable variation in the recovery of amphetamines (16-59%), opiates (33-50%), cocaine and benzoylecgonine (61-97%), carboxy-THC (0-53%) and PCP (9-56%). Less variation in collection volume, volume recovered and drug recovery was observed intra-device. The THC stability was evaluated in a common commercial collection protocol. Samples in the collection buffer were relatively stable for 6 weeks when stored frozen. However, stability was marginal under refrigerated conditions and poor at room temperature. Very little has been published on the efficacy of using IgG concentration, or any other endogenous marker, as a measure of OF specimen validity. Preliminary rinsing experiments with moderate (50 mL and 2 x 50 mL) volumes of water did not reduce the OF IgG concentration below proposed specimen validity criteria. In summary, obvious and more subtle variables in the OF collection may have pronounced effects on OF-drug concentrations. This has rarely been acknowledged in the literature, but should to be considered in OF drug testing, interpretation of OF-drug results and future research studies.