1 Possible cholinoceptor-mediated effects on lipolysis were investigated in vivo in human subcutaneous adipose tissue of non-obese, non-smoking, healthy subjects, by use of microdialysis. Cholinomimetic and sympathomimetic agents were added to the ingoing dialysate solvent. 2 Addition of nicotine to the perfusion solvent caused a concentration-dependent reversible increase in the levels of glycerol in the dialysate (lipolysis index). The opposite effect (also concentration-dependent and reversible) was caused by the addition of carbachol. The maximum effects were 100% stimulation and 50% inhibition, respectively, by nicotine and carbachol. Neither nicotine nor carbachol stimulated nutritive blood flow in adipose tissue (as measured with an ethanol escape technique). 3 The nicotine effect in situ was concentration-dependently counteracted by the nicotinic cholinoceptor antagonist, mecamylamine. Likewise, the carbachol effect was concentration-dependently counteracted by the muscarinic cholinoceptor antagonist, atropine. 4 When adipose tissue was pretreated with phentolamine plus propranolol in order to obtain a complete a and P-adrenoceptor blockade, the subsequent addition of nicotine or carbachol still induced an increase and decrease in dialysate glycerol levels (lipolytic or antilipolytic effects), respectively. When adipose tissue was pretreated with mecamylamine or atropine, the subsequent addition of acetylcholine caused a reversible decrease and increase, respectively, of the dialysate glycerol levels. 5 Nicotine and carbachol had no effects on glycerol release from human isolated subcutaneous fat cells that were incubated in vivo. 6 In conclusion, the data demonstrate a dual effect of the cholinoceptor system on glycerol output in human adipose tissue: stimulation through nicotinic receptors and inhibition through muscarinic receptors. These effects, which are not observed in vitro, are independent of the adrenergic system and the local blood flow and seem not to be mediated by a direct action on the fat cell.