Background Gastric emptying scintigraphy (GES) reports percent retention at 1, 2, and 4 h. Time to empty half the meal (T½) could simplify GE reporting. Aims To compare the performance of GES T½ to 1‐, 2‐, and 4‐h retention. Methods GES studies were reviewed; results determined according to retention at 1, 2, and 4 h. T½ was determined using 3 methods: (1) GES curve fitting using 0, 0.5, 1, 2, 3, and 4 h data; (2) linear interpolation using 0, 0.5, 1, 2, 3, and 4 h data; and (3) linear interpolation using only 0, 1, 2, and 4 h data. Results Of 495 patients, 265 had normal GE, 4 rapid GE (<30% retention at 1 h), and 226 delayed GE: 17 delayed only at 2 h (>60% ret); 94 delayed only at 4 h (>10% ret); and 115 delayed at both 2 h and 4 h. Strong correlations were seen between each T½ method and 1, 2, 3, and 4 h %‐empty values: curve‐fit T½ (r = −0.851, −0.942, −0.864, −0.744), linear T½ using all imaging times (r = −0.848, −0.972, −0.878, −0.763), and linear T½ using standard imaging times (r = −0.853, −0.974, −0.868, −0.760). The 132 min cutoff for delayed GE captures 99.1% to 100% of delayed GE at both 2 h and 4 h, 76.5% to 94.1% delayed at 2 h only, but only 36.7% to 39.4% delayed at 4 h only; 3.5 to 11.3% of patients with normal GE miscategorized as delayed. Conclusions GES T½ correlates more strongly with retention at 2 h than at 4 h. T½ alone may misclassify patients, particularly those with late‐phase (4 h only) delays, reducing its utility for diagnosing gastroparesis.