BACKGROUND Immediate hypersensitivity to corticosteroids is reported to occur with an incidence of 0.1%. The largest previous case series reporting corticosteroid skin testing has seven patients. METHODS We identified 23 patients (mean age 50 years, 65% female) from Auckland City Hospital who underwent skin testing (ST) for suspected corticosteroid hypersensitivity between July 2005 and April 2012. We performed a retrospective clinical case note review detailing clinical history of reaction, skin test results and subsequent management. Most patients (21/23) had a standard panel of testing with prednisolone, triamcinolone, methylprednisolone, hydrocortisone and dexamethasone. Skin tests used a 10% steroid stock concentration for skin prick tests (SPT) and dilutions of 1 : 1000, 1 : 100 and 1 : 10 for subsequent intradermal testing. A weal 3 mm greater than the negative control was considered positive. RESULTS A total of 23 patients were identified who had skin testing for suspected acute hypersensitivity to corticosteroids, eight of which had a history of anaphylaxis. From 28 reactions (in 23 patients), the most common route of administration was intra-articular (13), followed by oral (7), intravenous (3) and other (5). Skin tests were positive in 8/23 patients, and 7/8 of these patients had a history of corticosteroid-associated anaphylaxis. Skin tests were positive at either the skin prick test or intradermal stages. There was evidence suggesting clinical and skin test cross-reactivity between corticosteroids in one patient. One patient had a positive skin test, but negative oral challenge suggesting the skin test was false positive. Skin tests were negative in 15/23 patients. One patient had a negative prednisolone skin test and positive unblinded oral challenge, suggesting a false-negative skin test. CONCLUSIONS Skin testing can provide sufficient evidence to diagnose allergy in patients with a clear history of immediate hypersensitivity to corticosteroids such as anaphylaxis. Both skin prick and intradermal tests should be used. There is evidence of cross-reactivity between steroids, so a panel is recommended. False-positive and false-negative reactions do occur; however, the frequency is unknown. Challenge remains the only definitive way to demonstrate a safe alternative to use. CONCLUSIONS As the largest case series described, this article provides new evidence for the interpretation of skin tests when investigating possible immediate hypersensitivity to corticosteroids.