Changes in lipid levels/profiles can reflect health status and diseases. Urinary lipidomics, thus, has a great potential in clinical diagnostics/prognostics. Previously, only chloroform and methanol were used for extracting lipids from the urine. The present study aimed to optimize lipid extraction and examine differential lipid classes obtained by various extraction protocols. Urine samples were collected from eight healthy individuals and then pooled. Lipids were extracted by six solvent protocols, including (i) chloroform/methanol (1:1, v/v), (ii) chloroform/methanol (2:1, v/v), (iii) hexane/isopropanol (3:2, v/v), (iv) chloroform, (v) diethyl ether, and (vi) hexane. Lipid profiles of the six extracts were acquired by MALDI-TOF mass spectrometry (MS) and some lipid classes were verified by LIFT-TOF/TOF MS/MS. The data revealed that phosphatidylglycerol (PG) and phosphatidylinositol (PI) could be detected by all six protocols. However, phosphatidylcholine (PC) and sphingomyelin (SM) were detectable only by protocols (i)-(iv), whereas phosphatidylserine (PS) was detectable only by protocols (iii)-(vi), and phosphatidylethanolamine (PE) was detectable only by protocols (v)-(vi). In summary, we have demonstrated differential lipidome profiles yielded by different extraction protocols. These data can serve as an important source for selection of an appropriate extraction method for further highly focused studies on particular lipid classes in the human urine.