The objective of this study was to determine the incidence of hypomagnesemia in injured patients (versus a general hospital cohort) and to compare total and ionized values with each other and with the incidence of injury severity and ethanol level. It was a descriptive study of consecutive injured patients at a level II trauma center. For 3 months subjects underwent admission paired analysis of blood total magnesium (TMg) and ionized magnesium (IMg). IMg was determined by microanalysis of whole blood (Nova Biomedical, NovaSP9). During the same time period, all hospital samples for TMg (m = 1308) underwent simultaneous IMg testing. Pearson correlation coefficients were determined for IMg versus TMg, Injury Severity Score (ISS), Trauma Score (TS), Glasgow Coma Scale (GCS), and blood units transfused. By convention, hypomagnesemia was defined as TMg < or =1.6 mg/dl and IMg < or = 0.5 mg/dl. Altogether 43% of 113 trauma patients had low magnesium levels compared to 30% of noninjured cohorts (p<0.05). The correlation coefficient (r(2)) for TMg and IMg was 0.74 for TMg values >1.6, but for TMg < or =1.6 the r(2) was 0.35. Coefficients for IMg and ISS, TS, GCS, units transfused, and ethanol level were 0.06, 0.08, 0.10, 0.04, and 0.01, respectively. Mean IMg was 0.57+/-0.09 mg/dl with ethanol ingestion and 0.56+/-0.06 mg/dl without ethanol ingestion (mean +/- SD, p>0.05). It was concluded that hypomagnesia is common in injured patients but does not correlate with indices of injury severity or ethanol level. TMg is not a good predictor of IMg at low levels. Trauma patients may benefit from determination of IMg for accurate diagnosis of a low Mg level to facilitate repletion.