Objectives Colloid oncotic pressure is an important factor in cardiac surgery, owing to its role in haemodilution. The effect of cardiopulmonary bypass prime fluids on the colloid oncotic pressure are unknown. In this study, the effect of crystalloid and colloid prime fluids, with or without retrograde autologous priming on the colloid oncotic pressure during elective cardiac surgery were evaluated. Methods Randomized controlled trials and prospective clinical trials comparing crystalloid and colloid priming fluids or with retrograde autologous priming were selected. Primary outcome was the colloid oncotic pressure; secondary outcomes were fluid balance, fluid requirements, weight gain, blood loss, platelet count, and transfusion requirements. Results From 1582 records, 29 eligible studies were identified. Colloid oncotic pressures were comparable between gelofusine and hydroxyethyl starch during bypass (mean difference [MD]: 0.69; 95% confidence interval [CI]: -2.05, 3.43; P = 0.621), after bypass (MD: -0.11; 95% CI: -2.54, 2.32; P = 0.930), and postoperative (MD: -0.61; 95% CI: -1.60, 0.38; P = 0.228). Fluid balance was lower with hydroxyethyl starch than with crystalloids. Retrograde autologous priming reduced transfusion requirements compared with crystalloids. Blood loss was comparable between groups. Conclusions Colloid oncotic pressures did not differ between crystalloids and colloids. As a result of increased transcapillary fluid movement, fluid balance was lower with hydroxyethyl starch than with crystalloids. Haematocrit and transfusion requirements were comparable between groups. However, the latter was lower when retrograde autologous priming was applied to crystalloid priming compared with crystalloids alone. Finally, no differences in blood loss were observed between the groups.