Quantification of European Union (EU)-priority polycyclic aromatic hydrocarbons (PAHs) in plant matrices is a crucial task. Various methods for enrichment and preconcentration, such as the preloaded-pipette tip solid‑phase extraction (SPE) (1), are available. Nevertheless, analyte recovery as a result of homogenization, sample preparation, and extraction are rarely discussed in the field of phytopharmacy. This study deals with the recovery in dry plant extracts, which are typically used in phytopharmaceuticals and reflect the actual polycyclic aromatic hydrocarbon content in the commercially available end product (2). The aim of this study was to monitor benzo[a]pyrene, benzo[a]anthracene, chrysene, and benzo[b]fluoranthene loss of spiked samples as a result of commonly-used sample pretreatment, extraction, filtering, and evaporating techniques in 1:1 (v/v) cyclohexane–ethyl acetate primulae flos and sambuci flos dry extracts. Results showed that improper sample preparation can lead to false results. In the case of benzo[a]pyrene with a deviation of 155% from the theoretical true value.