Breast milk analysis provides useful information about acute newborn exposure to harmful substances, such as psychoactive drugs abused by a nursing mother. Since breast milk represents a complex matrix with large amounts of interfering compounds, a comprehensive sample pre-treatment is necessary. This work focuses on determination of amphetamines and synthetic cathinones in human breast milk by microextraction techniques (liquid-phase microextraction and electromembrane extraction), and their comparison to more conventional treatment methods (protein precipitation, liquid-liquid extraction, and salting-out assisted liquid-liquid extraction). The aim of this work was to optimize and validate all the extraction procedures and thoroughly assess their advantages and disadvantages with special regard to their routine clinical use. The applicability of the extractions was further verified by the analysis of six real samples collected from breastfeeding mothers suspected of amphetamine abuse. The membrane microextraction techniques turned out to be the most advantageous as they required low amounts of organic solvents but still provided efficient sample clean-up, excellent quantification limit (0.5 ng mL^-1), and good recovery (81–91% and 40–89% for electromembrane extraction and liquid-phase microextraction, respectively). The traditional liquid-liquid extraction as well as the salting-out assisted liquid-liquid extraction showed comparable recoveries (41–85% and 63–88%, respectively), but higher quantification limits (2.5 ng mL^-1 and 5 ng mL^-1, respectively). Moreover, these methods required multiple operating steps and were time consuming. Protein precipitation was fast and simple, but it demonstrated poor sample clean-up, low recovery (56–58%) and high quantification limit (5 ng mL^-1). Based on the overall results, microextraction methods can be considered promising candidates, even for routine laboratory use.