ALLERGIC reactions, including anaphylaxis, have been observed in dogs after vaccination (Brooks 1991, Greene 1998). Anaphylaxis is the most severe and life-threatening adverse event after vaccination. It is a systemic reaction of immediate type I hypersensitivity caused by allergen-specific immunoglobulin E, which causes the release of various mediators from mast cells and basophils (Greene 1998). In a previous retrospective study of 311 dogs in Japan with adverse reactions after vaccination, 46 of the dogs showed anaphylaxis, and seven of the 46 cases of anaphylaxis resulted in death (Ohmori and others 2002). Although allergic reactions, including anaphylaxis, after vaccination in dogs are a problem in small animal practice, little is known about them. The aim of this study was to obtain epidemiological information regarding allergic reactions after vaccination in dogs. Information on 85 dogs that had showed suspected allergic reactions within 24 hours after non-rabies vaccine injections was collected from 40 private small animal veterinary practitioners in Japan between January 2001 and July 2002; information on one to eight dogs (mean 2·1 dogs) was obtained from each practice. For the purposes of the study, the suspected allergic reactions were defined as clinical signs associated with type I hypersensitivity, such as hypotensive shock, dyspnoea, facial oedema, pruritus, weakness and diarrhoea, occurring after vaccination, as described previously by Greene (1998). The information on the 85 dogs comprised their history, the types of vaccine used, the history of previous vaccination, the clinical signs, and the time of onset of clinical signs after vaccination; the dogs were classified into three groups based on their clinical signs. Group 1 consisted of dogs with predominantly cardiovascular and/or respiratory signs, group 2 comprised those with predominantly dermatological signs, and group 3 consisted of dogs showing cardiovascular/respiratory and dermatological signs simultaneously. Of the 85 dogs, 83 were purebred and two were crossbred; there were 31 miniature dachshunds, seven shih tzus, six Maltese, six pugs, four labrador retrievers, four papillons, four Yorkshire terriers, three beagles, three miniature schnauzers, three Shetland sheepdogs, three toy poodles, two chihuahuas, two golden retrievers, two Welsh corgis and one akita, one Bernese mountain dog and one pomeranian. The dogs’ ages ranged from two months to 16 years; 43 were aged up to one year, 39 were over one year old, and the age of three of them was not clear. There were 42 males and 43 females. The following four types of vaccines are commercially available in Japan: a monovalent live parvovirus vaccine (type 1); monovalent inactivated parvovirus or Leptospira vaccines (type 2); combined live vaccines composed of canine parvovirus, distemper virus, adenovirus type 2 and/or parainfluenza virus (type 3); and combined live and inactivated vaccines composed of live canine parvovirus, distemper virus, adenovirus type 2, parainfluenza virus and/or coronavirus vaccines, and inactivated coronavirus and/or Leptospira vaccines (type 4). Of the 85 dogs studied, clinical signs of suspected allergic reactions were observed in two recipients of type 1 vaccine, 28 recipients of type 3 vaccines and 53 recipients of type 4 vaccines; no dogs had been injected with type 2 vaccines. Information on the types of vaccine used in two dogs was not obtained. Of the 85 dogs, 24 were classified into group 1: they showed predominant cardiovascular/respiratory signs consistent with anaphylaxis, such as circulatory collapse, cyanosis, bradycardia, hypotension, hypothermia, dyspnoea and/or tachypnoea. Of these 24 dogs, five showed concurrent gastrointestinal signs such as vomiting and/or diarrhoea. In addition, one dog, a seven-year-old male miniature dachshund, died of anaphylactic shock after vaccination. Fifty-nine of the 85 dogs showed dermatological signs such as facial oedema, erythema, pruritus and/or urticaria, and were classified into group 2; 55 of these dogs showed facial oedema, which was the most common dermatological sign. Of the 59 dogs in group 2, 11 dogs also showed concurrent gastrointestinal signs. Two of the 85 dogs showed cardiovascular/respiratory and dermatological signs at the same time (group 3), without concurrent gastrointestinal signs. The cardiovascular and/or respiratory signs in the 24 dogs in group 1 were observed within 60 minutes of vaccination, whereas the dermatological signs in the 59 dogs in group 2 appeared within 60 minutes in 22 dogs, but appeared from one to 24 hours after vaccination in the other 37 dogs (Table 1). Concurrent cardiovascular/respiratory and dermatological signs in the two dogs of group 3 were observed within 60 minutes after vaccination (Table 1). The results suggest that the cardiovascular/respiratory signs could have been caused by an immediate-type reaction, and the dermatological signs could have consisted of both immediateand non-immediate-type reactions, similar to the allergic reactions observed in human beings after vaccination (Kumagai and others 1997, Ohsaki and others 1999). Of the 85 dogs studied, 24 had received one vaccine injection, and 40 had received two or more vaccine injections, before the last vaccination which had provoked the suspected allergic reactions. Unexpectedly, 16 dogs had not received any vaccine injections before the vaccination which induced the reaction. In the other five dogs, the number of previous vaccinations was unknown. The 16 dogs with suspected allergic reactions after the first vaccination may have already been sensitised to some allergens included in the vaccines before the vaccination. The nature of such allergens in dogs is unclear, but may relate to the use of gelatin or bovine serum in the vaccines, as has been suggested in human beings (Bonin and others 1973, Erdos and others 1975, Kelso and others 1993, Sakaguchi and others 1995). It is also possible that anaphylactoid reactions induced the clinical signs in these 16 dogs, without sensitisation to allergens in the vaccines. In the present study, it was not possible to calculate the incidence of suspected allergic reactions because the total number of dogs that had received vaccination at the 40 veterinary practices was not obtained. Further studies, including the gathering of such information, will be necessary to invesShort Communications