Aqueous two-phase systems have been demonstrated to be a possible alternative to chromatographic separations during the industrial purification of proteins. While convenient high throughput screening methods were shown to drastically reduce experimental effort for the evaluations of ATPS as a unit operation, the selection of which phases to investigate is currently guided largely by prior knowledge. Correlations between protein descriptors and distribution were found, but the general applicability of such correlations especially under conditions of high protein load, is questionable, as currently no correlations take the saturation of the phases with protein into account. In this manuscript, we demonstrate how precipitation experiments using the phase forming components can guide the selection of both system type and tieline length for the purification of monoclonal antibodies. Phase selection and process development time can thus be significantly reduced, as all the necessary precipitation, binodal, and tieline experiments can be conducted within one day. Good qualitative correlations between precipitation data and both distribution and recovery of the target molecule were found. Most promising systems were selected for upscale to a 500 mL CPC. Influence of operation condition on the column and on HCP clearance was investigated. An increase in HCP clearance of more than threefold compared to batch extractions was observed. The importance of load protein concentration underlined the value of using a screening approach that incorporated target protein solubility data.