The invention discloses a method for calibrating an out-of-pile nuclear measurement ionization chamber. The method comprises the following steps: firstly, setting calibration conditions; when delta1(k) is greater than or equal to 5 percent FP or delta2(k) is greater than or equal to 3 percent FP, recalculating a calibration coefficient; then, calculating equivalent upper-section current IH(k) and lower-section current IB(k) of each channel and calculating out-of-pile axial deviation AOex(k); calculating relational expression coefficients a(k), b(k) and K(k) and calculating calibration coefficients alpha(k), KH(k) and KB(k) of each channel; identifying a result and outputting the result after requirements are met; calculating an out-of-pile nuclear measurement calibration coefficient by utilizing a linear relation between heat power and the sum of the equivalent upper-section and lower-section current of the ionization chamber and a linear relation between the out-of-pile axial deviation and in-pile axial deviation. By adopting the method disclosed by the invention, the defect that neutron flux distribution of a reactor core cannot be really reflected by out-of-pile nuclear measurement is overcome, and influences, caused by the sensitivity consumption of a detector and instrument drift, on a measuring result are reduced, so that the measurement of a neutron flux in the out-of-pile nuclear measurement accords with an actual neutron flux distribution condition better.