ObjectivesMagnesium (Mg) status, although difficult to assess, is suspected to be marginal in many individuals, especially athletes, and this has led to the common use of Mg supplements. The purpose of this article is to critique research that has addressed Mg supplementation in athletes.Data SourcesThe primary database was Medline, which was searched for English articles from 1966 to June 1999 using the words “magnesium” and “supplementation.”Study SelectionOnly experimental studies dealing with human subjects, Mg supplementation, and exercise performance were critiqued (n = 12).Data ExtractionQuality of critiqued articles was based on 1) use of cross-over designs, 2) how and if Mg status was assessed, 3) whether treatment was solely Mg supplementation, 4) duration of supplementation, 5) subject number, and 6) degree of experimental control. Articles were classified into “no effect” and “positive effect,” and also were examined in regard to the type of performance outcome (strength, anaerobic–lactacid, and aerobic).Data SynthesisMg is a cofactor to over 325 enzymatic reactions, and a deficiency of the mineral therefore has many physiological and exercise performance implications. Low dietary intakes, as found in many female athletes, coupled with increased urinary losses with exercise, may eventually lead to an Mg deficiency. Strength of evidence favors those studies finding no effect of Mg supplementation, regardless of whether the performance outcome was strength, anaerobic–lactacid, or aerobic. Analysis was confounded due to: 1) variable exercise modes, intensities, and durations, 2) variable training states and ages of subjects, 3) subject selection favoring males and gender differences has not been probed, 4) Mg dosage has ranged from 1 day to 3 months and from 116 mg/day to 500 mg/day, 5) multivitamins/minerals have been ingested with the Mg, 6) with one exception,1Mg status was either not reported or reliant on total Mg (TMg), 7) lack of a cross-over design, 8) only one study2made note of controlling exercise prior to exercise testing and blood assay, and 9) typical Mg intake measured was only measured in three of the studies.ConclusionsMost evidence indicates no effect of Mg supplementation on performance (strength, anaerobic–lactacid, and aerobic). When only peak treadmill speed during a VO2max test is examined, the strength of evidence is equivocal. Trained subjects appear to benefit less than untrained subjects, but this observation requires further study. Little research has focused on physically active females who may be at the highest risk for Mg deficiency. Research has been confounded by numerous factors.