This study was initiated to identify the signalment and clinical variables potentially associated with hypomagnesemia in horses evaluated at the North Carolina State University, College of Veterinary Medicine (NCSU-CVM) veterinary teaching hospital between January 1999 and May 2001. A nested case reference study (nested case-control study) was conducted to examine the potential relationship between hypomagnesemia and signalment, serum chemistry panel analyses, number of hospitalization days, discharge status, and diagnosis. A series of independent and multivariable logistic regression models were used to assess the potential association of each variable with low total serum magnesium concentrations. Four hundred one of 823 (48.7%) horses had serum total magnesium concentrations below the normal reference range. Hypomagnesemia was more likely to occur in horses older than I month of age. Colic (odds ratio [OR]: 2.96, 95% confidence intervals [CI]: 2.14-4.08), acute diarrhea (OR: 5.91, 95% CI: 2.32-15.06), other gastrointestinal disease (OR: 2.07, 95% CI: 1.15-3.71), infectious respiratory disease (OR: 5.07, 95% CI: 2.09-12.28), and multiorgan system disease (OR: 2.31, 95% CI: 1.24-4.28) were associated with hypomagnesemia in adult horses, whereas foals with diarrhea (excluding septic foals) (OR: 0.11, 95% CI: 0.01-0.84) were less likely to have hypomagnesemia. Overall, there was no relationship between hypomagnesemia and mortality (OR: 1.00, 95% CI: 0.72-1.41), but horses with colic and hypomagnesemia were less likely to die than horses with colic and normal or high total magnesium (OR: 0.53, 95% CI: 0.30-0.95). Among horses that survived, hypomagnesemia at admission was associated with a longer hospitalization period (OR: 1.45, 95% CI: 1.00-2.11).