In order to establish a soil diagnosis to avoid the circulation of contaminated pumpkin, the availability of heptachlor exo-epoxide (HEPX) in soil for uptake by pumpkin was evaluated. Namely, the relationship between HEPX concentrations in soils extracted with 50% (v/v) methanol–water solution and HEPX concentrations in pumpkin was investigated. We first tested the early-growth shoots of pumpkin (Cucurbita maxima Dutch. cv. Ebisu) planted in HEPX-contaminated soil in an environmentally controlled chamber. Significant (P＜0.01) relationships were found in regression analysis, which indicated that 50% (v/v) methanol–water solution is suitable for evaluating HEPX availability in soil by early-growth shoots of pumpkin. Next, HEPX concentrations in soils collected from the plant bases and extracted with 50% (v/v) methanol–water solution, and in the fruits of three varieties of pumpkin (C. maxima Dutch. cv. Ebisu, C. maxima Dutch. cv. Ajihei, and C. maxima Dutch. cv. TC2A) planted in eight fields, showed significant (P＜0.01) relationships. However, the coefficients of determination (R²) of the regression lines for pumpkin fruit were lower than those for the early-growth pumpkin shoots. This implied that the HEPX concentration in soil extracted with 50% (v/v) methanol–water solution, nodal positions of fruits, and dry matter production of fruits contribute to the HEPX concentration in pumpkin fruit. The result of multiple regression analysis using these factors showed that the adjusted R² values in all varieties increased compared to those determined by single regression analysis. However, from the viewpoint of soil diagnosis, it is better to estimate the HEPX concentration in pumpkin fruits not by using multiple factors, but by using the HEPX concentration in soil extracted with 50% (v/v) methanol–water solution only.