A method based on direct injection of diluted urine for the identification and quantification of amphetamine, methamphetamine, 3,4-methylenedioxymetamphetamine and 3,4-methylenedioxyamphetamine in human urine by electrospray ionisation liquid chromatography-tandem mass spectrometry was validated for use as a confirmation procedure in urine drug testing. Two deuterium labelled analogues, amphetamine-D5 and 3,4-methylenedioxymetamphetamine-D5, were used as internal standards. Twenty microliter aliquots of urine were mixed with 80 microL internal standard solution in autosampler vials and 10 microL was injected. The chromatographic system consisted of a 2.0 mmx100 mm C18 column and the gradient elution buffers used acetonitrile and 25 mmol/L formic acid. Two product ions produced from the protonated molecules were monitored in the selected reaction monitoring mode. The intra- and inter-assay variability (coefficient of variation) was between 5 and 16% for all analytes at 200 and 6000 ng/mL levels. Ion suppression occurred early after injection but did not affect the identification and quantification of the analytes in authentic urine samples. The method was further validated by comparison with a reference gas chromatographic-mass spectrometric method using 479 authentic urine samples. The two methods agreed almost completely (99.8%) regarding identified analytes when applying a 150 ng/mL reporting limit. Four deviating results were observed for 3,4-methylenedioxymethamphetamine and this was due to uncertainty in quantification around the reporting limit. For the quantitative results the slope of the regression lines were between 0.9769 and 1.0146, with correlation coefficients>0.9339. We conclude that the presented liquid chromatographic-tandem mass spectrometric method is robust and reliable, and suitable for use as a confirmation method in urine drug testing for amphetamines.