To study the effect of age on serum myoglobin more clearly, the analytical, intra-individual and inter-individual components of variation were estimated from duplicate analyses of specimens collected from 18 healthy elderly subjects [ages 74-97 years; 9 men (EM)], and 14 healthy younger subjects [ages 25-31 years; 7 men (YM)] over a period of 6 weeks. The mean values (microgram/L) were EM: 53.7; EW: 44.9; YM: 34.2; YW:24.8. Estimated analytical (CVA), intra- (CVI) and inter-individual (CVG) variations as CV% were: CVA: 2.2. CVI: EM: 13; EW: 9.9; YM: 12.4; YW: 9.6. CVG: EM: 37.6; EW: 28; YM: 18.5; YW: 13.4. The data obtained were used to derive the desirable analytical goal for imprecision (i.e., < or = 6.5% in EM; < or = 4.9% in EW and < or = 6.2% in YM; < or = 4.8% in YW); inaccuracy (i.e., < or = 9.9% in EM; < or = 7.7 in EW and < or = 5.5% in YM; < or = 4.12% in YW); the change required for serial results to be significantly different (i.e., 36% in EM; 28% in EW and 34% in YM; 27.2% in YW), the numbers of specimen collections required to produce a more precise estimate of the homeostatic set point of an individual within 5% (i.e., 26 in EM; 16 in EW and 24 in YM; 15 in YW), and the index of individuality (i.e., 0.34 in EM; 0.35 in EW and 0.67 in YM; 0.71 in YW). This study shows that intra-individual biological variation of myoglobin in healthy elderly subjects is not different from that in young subjects. Inter-individual variation, instead, is greatly influenced by differences in age and sex.