In　this　paper,　stress　distribution　in　patella　was　studies　by　3-D　photoelastic　experiments　and　finite　element　calculations.　The　experimental　conditions　were　in　good　agreement　with　that　of　body　mechanics:　(1)　The　ratios　of　m.　quadrieeps　femoris　force　FQ　to　lig.　patella　force　FP　were　1.03-1.42.　(2)　The　angles　between　FQ　and　FP　were　8.4　°　-18.7　°.　(3)　The　ratio　of　modulus　of　elasticity　for　epoxy　resin　and　silicone　rubber　matched　that　for　patella　and　cartilage,　i.e.,　145:1.　The　principal　stresses　and　their　path　line,　normal　stresses,　contact　stresses　between　patella　and　ossis　femoris　were　determined　in　various　flexion　angles　(15　°,　30　°,　45　°,　60　°,　75　°,　90　°).　Two　of　the　correlations　were:　(1)　Maximum　principal　stress　in　the　front　of　patella　(equation　presented)　(2)　Maximum　contact　stress　between　patella　and　ossis　femoris　(equation　presented)　where　W　was　body　weight　subjected　by　a　single　foot.　It　was　proved　that　the　transverse　break　of　patella　resulted　from　the　case　that　the　principal　stress　in　the　front　of　patella　exceeded　ultimate　tensile　strength.　And　Tension　Band　Wiring　(TBW)　therapy　conformed　to　the　law　of　physiological　stress　field.　A　modified　K-needles　position　in　TBW　therapy　was　suggested.　These　results　are　useful　for　further　research　of　human　patella　mechanics　and　recovery　therapy.　No　similar　result　has　been　found　in　American　MED　LIN　Data　Bank.　©　1991　SPIE.　All　rights　reserved.