I need help fitting a family of association kinetic curves.

The explanation on page 236-237 of Fitting Models to Biological Data Using Linear and Nonlinear Regression is not very complete. If you fit the model in the book, share Kon and Koff, but don't share Ymax (the maximum binding at a particular concentration of radioligand) since it willl depend on ligand concentration. Better, use the revised model below. In this model Ymax is computed from Bmax (maximum binding extrapolated to infinite radioligand concentration) and Kd (which is computed from Kon and Koff, which you are fitting). Share the value of all three parameters that need fitting: Kon, Koff and Bmax.

  
  <A>Radioligand=1.0e-9
  <B>Radioligand=3.0e-9
  Kob=Radioligand*Kon + Koff
  Kd=Koff/Kon
  Ymax=Radioligand*Bmax/(Radioligand+Kd)
  Y=Ymax*(1-exp(-1*Kob*X)

Rather than define 
radioligand within the equation, you can leave it as a parameter and then (in 
the Constraints tab) tell Prism to make the parameter Radioligand a dataset 
constant, and so get its value for each data set from the column title in the 
data table. In this case, the equation would look like: 
  Kob=Radioligand*Kon + Koff
  Kd=Koff/Kon 
  Ymax=Radioligand*Bmax/(Radioligand+Kd
  Y=Ymax*(1-exp(-1*Kob*X)