This study tested the possibility of adrenal autotransplantation in rats. Since the cortex and the medulla of the adrenal gland were from different origin embryologically, either whole adrenal glands (ADR), or capsule and cortex (CAP) or medulla (MED) were autotransplanted in the subcutaneous tissue. The functions of regenerated adrenal nodules were tested by measuring plasma corticosterone levels every fortnight. At the end of 9 weeks the rats were exposed to hypovolemic shock followed by naloxone injection to reverse the shock response. Results showed that rats transplanted with either cortex or whole adrenal started secreting corticosterone at 5 weeks post-transplantation (107.73 +/- 21.98 ng/ml, 126.04 +/- 48.41 ng/ml, respectively). Corticosterone levels increased to the value which were not significantly different from control by 9 weeks post-transplantation. However, rats transplanted with adrenal medulla showed very low corticosterone levels. Nine weeks post-transplantation, the mean blood pressure (MBP) of the CAP group was 135 +/- 13 mmHg and was not significantly different from sham-operated controls, whereas MBP of MED group was significantly lower than sham-operated animals (99 +/- 11 mmHg versus 141 +/- 9 mmHg). The MBP of the ADR group was also lower compared to sham-operated controls (112 +/- 17 mmHg P < 0.05). The MBP of the adrenal group was not statistically significant compared to the CAP group. After 1% body weight haemorrhage, the MBP decreased significantly in ADR (45 +/- 5 mmHg, P < 0.05) and MED group (36 +/- 9 mmHg, P < 0.001) compared to sham-operated rats (78 +/- 11 mmHg) but not in the CAP (56 +/- 9 mmHg). It was concluded that autotransplanted whole adrenal or adrenocortical tissues survived subcutaneously and produced sufficient corticosterone to alleviate haemorrhagic shock. Adrenal medullary tissue failed to regenerate subcutaneously and the presence of adrenal medullary tissue may suppressed the growth of transplanted adrenal gland.