Estimates are made of the error involved in the usual deuteron stripping calculation, which proceeds by means of the `zero-range approximation'. The finite-range effect may be simulated by maintaining the form of the zero-range calculation, but by multiplying the wave function of the captured neutron by a simple factor involving the optical-model potentials of the proton and deuteron, the potential describing the neutron binding and the binding energy of the deuteron. Numerical calculations, referring to the reactons 24Mg(d, p)25Mg, and 52Cr(d, p)53Cr, show that the effect on the angular distribution is very small.