Controlled　gradient　drilling　technology　was　put　forward　to　solve　the　problem　of　narrow　pressure　margins　in　deep　water　drilling.　In　this　paper,　an　optimization　model　for　optimal　drilling　parameters　design　is　established　for　controlled　gradient　drilling.　The　established　optimization　model　aims　at　achieving　the　longest　single　drilling　depth,　while　reducing　pressure　difference　between　wellbore　pressure　and　formation　pore　pressure　as　much　as　possible.　It　is　beneficial　to　increase　the　ROP　and　reduce　formation　damage.　The　SQP　(sequential　quadratic　programming)　algorithm　is　presented　to　solve　the　optimization　model.　Simulation　results　indicate　that　longer　single　drilling　depth　and　better　drilling　parameters　(control　variables)　can　be　obtained　by　use　of　optimization　model　when　the　range　of　control　variables　widens.　Furthermore,　in　case　study,　small　bottom-hole　pressure　difference　can　be　obtained　by　use　of　optimization　model.　In　addition,　the　established　optimization　model　ensured　the　dynamic　wellbore　pressure　within　narrow　pressure　margins　during　optimization　process.　It　is　conducive　to　safe　and　efficient　drilling.　©　Springer　Nature　Switzerland　AG　2020.