BACKGROUND: Skin prick tests of native spices (commercial powdered spices) are common in patients with allergy to birch or mugwort pollen. Clinical symptoms from spices are infrequent but occasionally severe. OBJECTIVE: To compare the skin prick test results with native spices and spice extracts and to determine the clinical relevance of test material. METHODS: Skin prick tests with the native spices coriander, caraway, paprika, cayenne, mustard, and white pepper were made twice at 2-month to 2.9-year intervals in 49 patients. During the latter time, tests were also made with spice extracts and spice-specific serum IgE was measured. RESULTS: The reproducibility of skin test results with native spices was 67% to 100%. Spice extracts, except white pepper, elicited positive skin test reactions in half those with positive reactions to native spices. Higher specific IgE concentrations (> or = 3.5 PRU/mL) were seen in cases where the skin tests were positive to the corresponding spices with 5% extracts of > 8 kD Mw. Three-fourths of the patients with positive skin tests to native spices were positive to birch pollen and one-half to a vegetable. Mild clinical symptoms from spices were reported by one-third. CONCLUSIONS: Spice allergens partly crossreact with those of pollens and vegetables. A minority of spice allergens may give clinical symptoms. The > 8-kD 5% extracts may be relevant skin prick test materials for identifying patients at risk of developing severe symptoms from ingested spices.