Cellular immunologic tests have not been used for diagnostic purposes in individuals at risk for autoimmune insulitis or in patients with partial beta-cell destruction because of a lack of studies that show their predictive value. In this study we initially evaluated 43 patients with recent-onset Type 1 diabetes (disease duration <or= 6 months, 29 ICA positive) with regard to beta-cell secretion stimulation test with glucagon and immunologic parameters, including CD4 +, CD8 +, CD4 + CD25 +, CD8 + CD25 + lymphocyte subsets. At baseline, C-peptide concentration 6 min after stimulation increased on average by 0.18 +/- 0.27 micro g/ml. The percentage of CD4 + cells was 42 +/- 9,4 % (healthy controls 44 +/- 7.3 %, p nonsig.) and percentage of CD8 + was 33 +/- 8.6 % (healthy control 31 +/- 8.3 %, p nonsig.). Relative size of CD4 + CD25 + subpopulation was 7 +/- 5.4 % (healthy control 2 +/- 2 %, p < 0.001). Percentage of activated CD8 + cell subset was also increased (2 +/- 1.4 vs. 1.0 +/- 1.0 %), but not significantly. Functional beta-cell testing was repeated after 6 months and nineteen patients were eligible for analysis. Their response was weaker after 6 months (0.13 +/- 0.1 micro g/ml, p < 0.05 vs. baseline). The average change in C-peptide excursion from baseline to the endpoint was - 0.07 +/- 0.17 micro g/ml. There was no significant correlation between beta-cell functional parameters at baseline (C-peptide6min (baseline)) and the relative size of various T cell subpopulations. Results were identical for the 6-month beta-cell functional data (C-peptide6min (6month)). The change in the excursion of C-peptide between baseline and follow-up visit (C-peptide6min (6month-baseline)) showed mild, negative correlation with relative size of the CD8 + CD25 + subpopulation (r = - 0.511, p = 0.025), which may indicate that the size of this cell subpopulation has predictive value in assessing future functional beta-cell changes.