The work aims to explore the effect of different ice addition ratios on broccoli quality during ice-assisted transport and develop a quality prediction model. The fresh broccoli was packed in foam boxes with varying ice-to-broccoli ratios (1∶1.5 and 1∶2.5) and stored at 20 ℃ to simulate transport conditions and then the key quality indicators were regularly measured. Additionally, a grey model was constructed to predict the quality by combining the O2 content in the microenvironment in the foam box with chlorophyll and mass loss rate. Compared with the Ice1.5 group (1∶1.5), Ice2.5 group (1∶2.5) could delay the yellowing rate and the increase in weight loss rate of broccoli and inhibit the decline rate of chlorophyll content and respiratory intensity, having better preservation effect. The GM(1,2) grey model effectively predicted chlorophyll content with minimal error. In summary, Ice2.5 is more beneficial for maintaining broccoli quality during ice-assisted transport and the GM(1,2) grey model can realize low-cost prediction of broccoli quality and provide theoretical guidance for quality monitoring of broccoli in ice-assisted transport.