In this work, laboratory-made green beers were produced using a commercial hopped malt extract and used to study their cross-flow microfiltration (CFMF) performance in a bench-top plant, appropriately designed and equipped with ceramic tubular membrane modules with nominal pore size of 0.4, 0.8, or 1.2 μm, under different values of the initial green beer turbidity (H ≡ 0.7–61.9 European Brewery Convention (EBC) unit), feed superficial velocity (v S ≡ 2–6 m s−1), and transmembrane pressure difference (TMPD ≡ 0.97–4.73 bar) under constant temperature (~10 °C). Whatever the membrane pore size and rough beer turbidity, the minimum quasi-steady-state value of the overall membrane resistance $$ \left({\mathrm{R}}_{\mathrm{T}}^{*}\right) $$ was obtained at TMPD of 3 to 4 bar and v S = 6 m s−1. The 0.8-μm pore membrane was selected for the loss in the permeate density, and color was limited, while the volumetric permeation flux was high enough. Then, it was used to assess a double logarithmic trend between the quasi-steady-state permeation flux (J*) and beer turbidity with a limiting flux of 63 ± 6 L m−2 h−1 for H > 21 EBC units. Precentrifuged beer feeding resulted in 2.7- to 4.1-fold increase in J* with respect to the above limiting flux, provided that the haze level of rough beer had been reduced to 1.0 or 0.7 EBC unit, respectively. The cake filtration fouling mechanism was identified as the main reason for flux decline by analyzing mathematically the time course of both the experimental permeate volume and permeation flux data. By operating with the aforementioned clarified green beers at TMPD of 3.73 bar, v S of 6 m s−1, and periodic CO2 backflushing, the average permeation flux (300–385 L m−2 h−1) resulted to be from three to five times higher than that (80–100 L m−2 h−1) at 0–2 °C claimed by the three commercial CFMF processes currently available.