PRIMARY hypoadrenocorticism is a common endocrinopathy in dogs. Hypercalcaemia is recognised in approximately 30 per cent of clinical cases of hypoadrenocorticism (Peterson and Feinman 1982) and in adrenalectomised dogs (Rogoff and Stewart 1928). These studies have measured total calcium only and it is unclear whether or not the hypercalcaemia represents an increase in the biologically active ionised calcium fraction. The mechanism of hypercalcaemia is not fully understood; a number of possible causes have been suggested including volume contraction, increased intestinal absorption, increased parathyroid activity and decreased urinary excretion (Walser and others 1963, Wajchenberg and others 1965, Walker and Davies 1981, Peterson and Feinman 1982). It has also been suggested that hypercalcaemia is associated with more severe sodium and potassium abnormalities (Peterson and Feinman 1982). This short communication provides the results of an investigation into electrolyte disturbances in 36 dogs with hypoadrenocorticism. Medical records at the Queen Mother Hospital for Animals, Royal Veterinary College were reviewed retrospectively and 36 dogs with confirmed hypoadrenocorticism and complete biochemistry results were identified. Dogs that had received glucocorticoid therapy in the past 14 days were excluded. Hypoadrenocorticism was confirmed using an adrenocorticotrophic hormone stimulation test (Melian and Peterson 1996). Signalment, initial blood results (packed-cell volume [PCV], albumin, sodium, potassium, total calcium, ionised calcium, pH and creatinine), and total and ionised calcium values following treatment were recorded. Electrolyte and acid base values were measured using an amperometric autoanalyser (CCX; NOVA biomedical) on anaerobically handled heparinised blood within one to two minutes of venipuncture. A reference range for the autoanalyser had been previously established using 60 normal dogs. The sodium: potassium ratio (Na: K) was calculated. Abnormal calcium levels were classified as values outside the reference range in total calcium, ionised calcium or both. Descriptive statistics and correlation analyses for the data were generated using a statistical software package (SPSS 14· 0). Data were analysed for associations between total and ionised calcium and measured variables (albumin, pH, PCV, sodium potassium, creatinine and Na: K). These variables were chosen to identify both markers of severity (creatinine and Na: K) and possible causes of hypercalcaemia (for example, pH and creatinine, PCV and albumin as measures of volume contraction). Pearson’s and Spearman’s correlations were used to explore these relationships in parametric and non-parametric data, respectively (Harris and Taylor 2004). Ninety-four per cent of dogs were pure breeds with a number of different breeds being represented. Twenty one (59 per cent) were neutered females, one (3 per cent) was an entire female, 11 (30 per cent) were castrated males and three (8 per cent) were entire males. The mean (sd) age at presentation was 5· 1 (2· 6) years. These proportions are similar to previously described populations of dogs with hypoadrenocorticism. Full blood gas and ionised calcium were available in 31 cases. Fifty-two per cent (16) of the dogs had abnormal ionised calcium (42 per cent increased; 10 per cent decreased). Total calcium was normal in 19 per cent (three) of these. Abnormalities in total calcium were present in 47 per cent (17) cases (42 per cent increased; 5 per cent decreased). Concurrent abnormalities in total and ionised calcium were present in 22 per cent (eight) cases; one …