As previously shown, 14 mM D-glucose, a non-insulinotropic concentration in isolated chicken pancreas, permits an insulin release in response to D-glyceraldehyde, (D-GA; a glycolytic fuel) and L-leucine or alpha-ketoisocaproic acid (alpha-KIC) (non-glycolytic fuels), which alone are not initiators of insulin release in this species. 2. The "permissive" effect of D-glucose was also observed in the presence of D-mannose (which, as shown herein, is not insulinotropic alone). 3. The specificity of glucose for this "permissive" effect was, therefore, subsequently questioned in the presence of 10 mM alpha-KIC by substituting various glycolytic and non-glycolytic fuels to glucose. 4. D-GA (at 5 and 15 mM), D-mannose (30 and 50 mM), or the association of L-glutamine + L-asparagine permitted an insulin release in response to alpha-KIC. 5. The response was, however, delayed with D-GA, only occasionally with 50 mM D-mannose, and required high concentrations and was delayed in the presence of L-glutamine + L-asparagine as compared to that obtained with 14 mM D-glucose + alpha-KIC. 6. In conclusion, the threshold of fuel-induced insulin release is much higher in the chicken than in mammals and this threshold is most efficiently lowered by glucose.