There are scant data from African populations on the association between beta-cell function and response to treatment with oral hypoglycaemic agents in Type 2 diabetes mellitus (T2DM). Fasting plasma C-peptide (FCP) and glucagon-stimulated C-peptide (GSCP) levels were measured in 116 Nigerians with T2DM at a university teaching hospital. After 9 months of follow-up and treatment, they were categorized into three groups based on response to treatment: (A) good control but not on maximum sulphonylurea (SU) therapy, (B) inadequate control but not on maximum SU therapy and (C) on maximum SU therapy+/-insulin or biguanide. Logistic regression models were used to investigate how well C-peptide levels predicted the subjects belonging to Group C who are likely to require insulin. The mean FCP and mean GSCP levels of Group C were significantly lower than in the other groups (p=0.024; p= <0.001 respectively). A GSCP cut-off value of < or =1.3 ng/mL predicted membership of Group C with 85% sensitivity and 89% specificity while a cut-off of < or =1.8 ng/mL was associated with 91% sensitivity and 66% specificity. In resource-poor settings where inadequate treatment are common, estimation of GSCP may be useful in predicting treatment response and should be weighed against the cost of inadequate therapy with higher morbidity and mortality.