Many clinical and forensic toxicology laboratories establish criteria for identifying a random urine specimen submitted for drug screening as being "normally concentrated" or "dilute" by incorporating creatinine analysis and/or specific gravity measurement into their testing protocols. The objective of this study is to describe the importance of urine creatinine analysis and specific gravity measurements in the Correctional Service of Canada (CSC) drug-testing program. The CSC program uses the Substance Abuse and Mental Health Services Administration (SAMHSA) creatinine cutoff value (20 mg/dL) mandated for workplace drug testing in the United States. In the CSC program, urine specimens must have a creatinine concentration <20 mg/dL and specific gravity value </=1.003 to be considered dilute. The CSC program also incorporates lower drug/drug metabolite screening and confirmation cutoff values (dilution protocol) for specimens that are administratively defined as dilute. Seventy-nine hundred and twelve urine specimens from 2000 to 2002 (6.8% of total workload) were defined as dilute. Twenty-six percent of all dilute specimens (n = 2054) screened positive for one or more drugs using the SAMHSA cutoff values. The screening-negative dilute specimens were taken through the dilution protocol scheme with lower screening cutoff values and confirmation cutoff concentrations at the lower limits of quantification (LLOQ) for each method. Over 1100 of the 5858 dilute urine specimens (18.8%) confirmed positive for one or more drugs in 2000 to 2002 when taken through the dilution protocol scheme. CSC workload is separated based on whether specimens are referred from institutions or from community settings (such as parolee programs). The positive rate for dilute specimens averaged 18.2% from CSC institutions and 22.3% from specimens collected from parolee specimens in 2000 to 2002. The drugs most often confirmed in dilute specimens from institutions were cannabinoids (annual positive rate ranged from 13.7 to 18%) and codeine and/or morphine (ranged from 0.2 to 2.8%). The drugs most often confirmed in dilute urine specimens from community settings in 2000-2002 were cannabinoids (annual positive rate ranged from 10.3 to 12.5%) and cocaine metabolite (ranged from 6.6 to 10.3%). In conclusion, one can reduce the false-negative rate for drugs of abuse in urine drug testing programs by incorporating lower screening and confirmation cutoff (eg, LLOQ) concentrations for dilute specimens that screen negative for drugs of abuse when using the SAMHSA mandated screening and confirmation cutoff concentrations.