Few studies have investigated factors influencing the efficacy of chemotherapy in older patients with cancer. This study aimed to evaluate the usefulness of G8, geriatric assessment (GA), and factors measured in general clinical practice for evaluating progression-free survival (PFS) of first-line palliative chemotherapy in older patients with advanced gastrointestinal cancer. This was a prospective observational study of older patients (age ≥70 years) with advanced gastrointestinal cancer. The modified cut-off value of G8 was determined by referring to two or more abnormal GA conditions. The usefulness of baseline GA and G8 (conventional and modified cut-off value) was assessed according to the efficacy (PFS and disease control rate) of the administered first-line palliative chemotherapy. Overall, 93 patients were evaluated between March 2017 and February 2019. A modified G8 cut-off value of ≤12 had a sensitivity and specificity of 68.9% and 46.9%, respectively. PFS was significantly prolonged in the patients with G8 > 12, serum albumin ≥3.5 g/dl, and in whom grade ≥3 adverse events occurred. There was no significant difference in the PFS between monotherapy and combination therapy. GA was not useful for predicting PFS prolongation or the occurrence of serious adverse events in first-line treatment. Among older patients with advanced gastrointestinal cancer who receive first-line chemotherapy, a modified G8 cut-off value of 12 points, occurrence of grade 3 or higher adverse events, albumin levels, rather than age or performance status were predictors of PFS prolongation.