The TEM-EDS bulk-tissue analysis procedure described involves: (1) low-temperature oxygen-plasma ashing of soft tissue specimens weighing greater than 20 micrograms in aluminum-foil crucibles; (2) the solubilization of the ash in 5 microliter 0.5N HNO3 containing a known quantity of cobalt as an internal, non-interfering reference element; (3) the spraying of the solubilized ash from glass microcapillary tubes onto thin carbon-collodion films mounted on titanium grids; (4) EDS analysis of individual microdroplets approximately 3 micron in diameter; (5) the quantitation of elemental concentrations from the element: cobalt intensity ratio by the "ratio model" technique. This technique was assessed and found to yield linear curves (greater than or equal to 0.999) for elements in 'artificial tissue' standards (concentration range = 5 - 340 mM kg-1 dry weight). The overall reproducibility of the technique is therefore quite good (e.g. error of 4.7% for P and K in 25 analyses) within the range of concentrations expected for most of the major biological elements encountered in vertebrate and invertebrate soft tissues. Absolute accuracy can be improved with quantitative procedures that account for peak-overlapping and escape peak contributions etc., so that the ultimate MDL for sodium may well be of the order of 1 mM kg-1 dry weight. The usefulness of the technique for the provision of basic biochemical information (especially in invertebrate systems which have received but meagre attention) is illustrated: (a) by comparing the calcium content of male and female blood-flukes (Schistosoma mansoni) in mixed-sex and unisexual laboratory infections; and (b) by determining the changes induced by daily injections of the drug Astiban on the element composition of female Schistosoma. We conclude that the technique can represent a useful multi-element detection facility which offers certain pertinent advantages over alternative microchemical techniques, such as atomic absorption spectrophotometry.