The aim of this cross-sectional study was to examine the number of days required to estimate habitual vegetable variety by conducting a multiday, dietary record. Sixty respondents from three groups in Japan (rural residents, general students, and nutrition students) participated in the study using a self-administered questionnaire in September 2018. To measure vegetable variety, the number of different vegetables consumed was extracted from the dietary records of seven consecutive days. Differences in the number of vegetables consumed and the capture proportion over seven consecutive days between groups were examined using repeated measures analysis of variance and one-way analysis of variance. The vegetable variety between each day was also compared using Pearson’s correlation coefficient. The vegetable variety based on dietary records for seven consecutive days confirmed the differences between groups by repeated measurements (p = 0.013). However, there was no significant difference among groups in the capture proportion per survey day based on seven consecutive days. Furthermore, there were significant correlations between the number of vegetables consumed over seven consecutive days and that consumed on two or more days (r > 0.50, p < 0.01) and especially three or more days in all groups (r > 0.70, p < 0.001). The present study suggested that a dietary survey over two or more days could provide an estimate of habitual vegetable variety.