The use of solid-phase filtration column (MycoSep 228®) and liquid chromatography for determination of patulin in apple and hawthorn beverages was investigated. The method passed a series of validation tests, including separation efficiency, sensitivity, repeatability, validation within a collaborative trial, recoveries, and applicability. Accuracy and precision were very good for determination of patulin in both spiked and naturally contaminated apple/hawthorn beverage samples at levels of 10 μg/L and above, with limits of detection and quantification at 8 and 23 μg/L, respectively. The applicability of the method was tested by analyzing commercially available domestic apple and hawthorn beverages. Patulin was found in 3 of 52 (5.77%) apple beverages at concentrations ranging from 63.45 to 88.44 μg/L, and in 6 of 43 (13.95%) hawthorn beverages ranging from 19.8 to 206.88 μg/L. Among these were 3 apple and 4 hawthorn beverage samples with toxin levels that exceeded the Chinese legislative requirement of 50 μg/L in apple- and hawthorn-based products intended for human consumption. Comparable tests revealed that the results obtained by both the new MycoSep-based method and the official AOAC method are not significantly different (t value 0.003268 < t0.05, P > 0.05).