A role for liver mitochondria in the regulation of cytoplasmic Ca’+ concentration was first suggested by their ability to accumulate large amounts of Ca’+ (Drahota et a/.. 1965) and to maintain extramitochondrial Ca2+ at 0 . 3 2 p ~ (Becker ct a/., 1980; Brand & De Selincourt, 1980; Nicholls, 1978), described by Nicholls (1978) as the set point which may be altered by varying the rates of Ca2+ uptake and efflux. This has led Joseph et al. ( 1982) and Crompton (1 985) to argue that liver mitochondria operate to buffer cytoplasmic Ca2+ in the intact cell. Liver mitochondria studied this way have contained more than 28 nmol of Ca2+ /mg of protein, e.g. Nicholls ( 1978), yet recent electron probe micro-analyses (Somlyo et a/., 1985) of liver cells and use of special fractionation techniques (Reinhart et a/., 1984) sugqest that mitochondria in situ contain less than 2 nmol of Ca +/mg of protein. Hansford & Castro (1982) showed lack of buffering by rat heart mitochondria when they contained less than 5nmol of Ca2+/mg of protein and concluded that heart mitochondria act to vary matrix Ca” and so operate in the mode of dehydrogenase control (Denton & McCormack, 1985). The purpose of the present work is to examine the behaviour of liver mitochondria with low, probably physiological, Ca2+ content towards external Ca2+ at concentrations found in hepatocyte cytoplasm. Mitochondria were prepared by the method of Chappell & Hansford ( 1972) in 250 mM-sucrose/ 10 mM-Hepes/l mMEGTA/pH 7.4, with the exception that the third centrifugation at 10000 rev./min was carried out in 250 mM-sucrose/ 5 mM-Hepes/Io mM-NTA/pH 7.4. The pellet was then resuspended in 50ml of the latter medium and left to stand on ice for 25--30 min before being centrifuged at 10 000 rev./min for 10 min. The resulting mitochondrial suspension contains 0.9 nmol of Ca” /mg of protein and 35 nmol of Mg2+/mg of protein corrected for medium Ca2+ and Mg2+ content. Ca2+ content was uninfluenced by alterations in pH (6. I-7.9), NTA concentration (1-17 mM), or by substituting 250mMsucrose with 125 mM-KCI. pCa was determined in mitochondrial incubations using a Ca” -sensitive electrode (Radiometer type F2002 with a KCI reference electrode Radiometer type K801) and calibrated using Ca-NTA buffers of known pCa (Martell & Smith. 1977; Fabiato & Fabiato, 1979). Medium Ca’+ content and mitochondrial total Ca2+ content were measured with a Perkin-Elmer atomic absorption spectrophotometer Model 380 after mitochondrial Ca” had been extracted with 5% (w/v) trichloroacetic acid/l.5% (w/v) LaCI, (final concentration) and the denatured protein removed by centrifugation. One may calculate the buffer capacity ( B ) for Ca2+ of mitochondria (B,,, ,) by comparing the rise in free external Ca” on addition of Ca2+ aliquots to 10mM-NTA plus mitochondria (B,,,,,) and on addition to IOmM-NTA alone ( B , m M N T A ) . Buffer capacity is measured by the amount of Ca + required to alter pCa by one unit. 4 -