Despite improvements in outcomes for human islet transplantation, characterization of islet preparations remains poorly defined. This study used both light microscopy (LM) and electron microscopy (EM) to characterize 33 islet preparations used for clinical transplants. EM allowed an accurate identification and quantification of cell types with measured cell number fractions (mean±s.e.m.) of 35.6±2.1% β-cells, 12.6±1.0% non-β-islet cells (48.3±2.6% total islet cells), 22.7±1.5% duct cells, and 25.3±1.8% acinar cells. Of the islet cells, 73.6±1.7% were β-cells. For comparison with the literature, estimates of cell number fraction, cell volume, and extracellular volume were combined to convert number fraction data to volume fractions applicable to cells, islets, and the entire preparation. The mathematical framework for this conversion was developed. By volume, β-cells were 86.5±1.1% of the total islet cell volume and 61.2±0.8% of intact islets (including the extracellular volume), which is similar to that of islets in the pancreas. Our estimates produced 1560±20 cells in an islet equivalent (volume of 150-μm diameter sphere), of which 1140±15 were β-cells. To test whether LM analysis of the same tissue samples could provide reasonable estimates of purity of the islet preparations, volume fraction of the islet tissue was measured on thin sections available from 27 of the clinical preparations by point counting morphometrics. Islet purity (islet volume fraction) of individual preparations determined by LM and EM analyses correlated linearly with excellent agreement (R²=0.95). However, islet purity by conventional dithizone staining was substantially higher with a 20-30% overestimation. Thus, both EM and LM provide accurate methods to determine the cell composition of human islet preparations and can help us understand many of the discrepancies of islet composition in the literature.