The interaction of fluorescein-labeled nerve growth factor (NGF) with human melanoma cells (A875) has been studied in order to assess better methodology for rigorous NGF binding studies. The NGF was modified at a single carboxyl group with iodoacetamidofluorescein after reaction with carbodiimide and cystamine. The modified NGF showed full binding competence in competition with radiolabeled NGF and full biological activity in neurite outgrowth assays compared to native NGF. Binding to unfixed, viable cells was assayed using flow cytometry. This method offers the advantage that unbound ligand need not be separated from that which is cell-associated, thus avoiding perturbation of the binding equilibrium, and accurate, extensive statistical analysis is possible. Binding of fluorescein-NGF was mainly specific and saturable, with analysis by three methods of data treatment indicating a Kd of 0.8 to 3 nM at 4 degrees C. Time-based data acquisition allowed a continuous time course for binding to be generated. Binding reached a steady-state level within 5 min of exposure of the cells to the ligand. Kinetic and steady-state results obtained using fluorescein-NGF agree well with previous data produced by 125I-NGF binding studies. The main limitation of the flow cytometric method in the NGF system is the relative lack of sensitivity compared to the binding of radiolabeled NGF, partially due to unusual quenching of the fluorophore bound to NGF.