[en] Applicability of the sample worth method has been investigated in the rod worth experiment made with FCA Assembly VII-1/90Z-7, mock-up of the prototype fast reactor. A B₄C pin cluster of B73-16P (73% enriched B₄C, ¹⁰B asymptotically equals 1.8 kg) was mainly used for the simulating rod. The correction factor for the rod interaction was measured using three kinds of samples, i.e. B90-1P (90% enriched B₄C, ¹⁰B asymptotically equals 0.14 kg), BN-16P (natural B₄C, ¹⁰B asymptotically equals 0.5 kg) and B73-16P. The reactivity worth of B90-1P sample is about 1/7 that of B73-16P. Agreement was good between the correction factors obtained with each sample, the experimental error being about 0.5 to 1%. The error tends to increase with decreasing reactivity worth of the sample used. The reactivity worth of the multiple rod worth estimated by using the correction factors and the single rod worths agreed very well with that obtained by direct measurement. The sample worth method is useful for estimating the reactivity worth of a multiple control rod. A sample having more than half the reactivity worth of the control rod is advisable for the experiment. (author)