It may be possible to use a mixture of known peptides run on a linear gradient to create a model that can be used to predict an unknown LC gradient of unknown samples based on deviation from the model. Data from RM 8321 can also be used to test other de novo search algorithms since it is a known mixture of 439 peptides. This data set consists of LC-MS/MS analysis of RM 8321 on a linear gradient of five varied lengths (30 min, 60 min, 90 min, 120 min, and 240 min), which can be used for modeling or testing de novo searching. To facilitate future testing of LC gradient model performance, separate E. coli protein digest (Waters) and HeLa protein digest (Pierce) were separated on a typical two-step gradient. Specific LC gradient separations are described within and should be consulted. All separations, except ?130 min?, were run with the same data acquisition settings, and all were analyzed using a Thermo Fusion Lumos tribrid mass spectrometer running data-dependent acquisition using the orbitrap detector for both MS1 and MS2. These instrument method files are also available and named obviously.