An analytical procedure was developed for the detection and quantitation of diazepam in cream biscuits, which were used to commit crime. The method involves the extraction of diazepam with ethanol at room temperature, and the extract is filtered, evaporated to dryness, and redissolved in the mobile phase, methanol-acetonitrile-tetrahydrofuran-water (15 + 55 + 4 + 26, v/v). The separation is achieved on a C18 reversed-phase column with the mobile phase and diode array detection (lambda(max)) at 230 nm. Medazepam is used as the internal standard is for quantification. The calibration plot for the determination of diazepam is based on linear regression analysis (y = 0.6687x + 0.0372; r2 = 0.995). The limit of detection for diazepam in the biscuit samples was estimated as 600 ng/mL. The limit of quantitation for diazepam was estimated as 1.75 microg/mL. The diazepam detected per piece of biscuit was found to be in the range of 0.27-0.45 mg. Pure diazepam was added to biscuit samples at 3 levels (100 and 500 microg/g, and 1 mg/g), and the recoveries were found to be 95%. The mean retention time of diazepam was 2.7 min and that of medazepam (IS) was 4 min. The relative standard deviations of the diazepam level in the biscuit samples were estimated to be 0.4% for retention time and 1.02% for peak area in intraday analysis, whereas the corresponding values were and 0.61 and 2.34% in interday analysis. The method is rapid and reliable for qualitative and quantitative analysis of cream biscuits laced with diazepam, and it can be used by law enforcement laboratories for routine analysis.