A sensitive and rapid method based on alcohol-assisted dispersive liquid–liquid microextraction followed by high-performance liquid chromatography for the determination of fluoxetine in human plasma and urine samples was developed. The effects of six parameters on the extraction recovery were investigated and optimized utilizing Plackett–Burman design and Box–Benken design, respectively. According to the Plackett–Burman design results, the volume of disperser solvent, extraction time, and stirring speed had no effect on the recovery of fluoxetine. The optimized conditions included a mixture of 172 L of 1-octanol as extraction solvent and 400 L of methanol as disperser solvent, pH of 11.3 and 0% w/v of salt in the sample solution. Replicating the experiment in optimized condition for five times, gave the average extraction recoveries equal to 90.15%. The detection limit of fluoxetine in human plasma was obtained 3 ng/mL, and the linearity was in the range of 10–1200 ng/mL. The corresponding values for human urine were 4.2 ng/mL with the linearity range from 10 to 2000 ng/mL. Relative standard deviations for intra and inter day extraction of fluoxetine were less than 7% in five measurements. The developed method was successfully applied for the determination of fluoxetine in human plasma and urine samples.
