A standard, dual-isotope meal (Tc-99m-egg sandwich and In-111 DTPA in water) was administered to 14 normal volunteers and 37 patients, who had not undergone gastric surgery, to determine if the emptying characteristics of the liquid phase alone could accurately predict delayed solid emptying. Delayed gastric emptying was defined clinically as a solid half-emptying time more than two standard deviations greater than the mean for normal volunteers. Linear regression analysis was performed on the natural logarithm of liquid fractional retention at each time interval to yield a slope and y-intercept for each subject. There was no significant difference (0.6 less than P less than 0.8) between volunteers and patients with normal solid emptying for the liquid slope. In patients who exhibited delayed gastric emptying for solids, the liquid slopes were significantly different from the normal values (P less than .001). There was a high correlation of liquid slope to solid half-emptying time in all patients and volunteers (r = -0.80, P less than .001). Comparison of the liquid slope measurement to solid half-emptying time criteria revealed a sensitivity of 96%, a specificity of 100%, and a predictive value of 100% for the slope test. These results suggest that delayed gastric emptying can be accurately detected with a liquid-solid meal using only a single-liquid label.