In a double blind study, alcohol/onion extract (5% ethanol) was injected simultaneously with 20 IU and 200 IU rabbit anti‐human‐IgE intradermally in 12 adult volunteers (6 atopics, 6 non‐atopics). Diameters of wheals and flares were measured 10 mm after and compared with control sites challenged with 20 IL1 and 200 IL1 anti‐IgE in a 5% ethanol solution. The skin sites were then treated epidermally with 45% alcohol/onion extract and 45% ethanol under occlusion. Diameters of late cutaneous reactions were measured hourly. Oedema formation was clinically estimated according to an arbitrary scale and skin thickness measured with a calliper. In the onion‐treated skin sites the wheal areas were significantly reduced (20 IU: control: 108 ± 53 mm2: onion 69 ± 42 mm2, P<0.05;, 200 IL anti‐lgE: control: 152 ±, 25 mm2, onion: 138 ±, 26 mm2, P<0.02). The oedema formation during the late phase skin reaction was markedh depressed (P<0.005 at 2 h, P<0.01 at 4 and 6 h, P<0.02 at 8 h). The extent of late skin reactions was slightly, but not significantly reduced. Obviously. onions contain pharmacologically active substances with anti‐inflammatory and or al