ABSTK4CT Shortly a/ier the death v,oLrld otcur the change.s that probubLy, e.tti tote tirne of the tleuth. One of the changes ttre a decreuse of post-nofiem tetnperuture as meosured by tanalometer lhrough rectLtnt. ,\lodiJicution oJ the nrcttsuremenl v'ith tligitul lo otoDret$ on the a:illa bccause of'rhe diJJicttlty to.find the tonatometer. Thi.s stutll, v'us contlucted to tletermine relationship on.l patterns ol decline in botly temperutu'e os measttred by digirul lonotomeler on the axilla to the upproximate length oJ time oJ deoth and ohtuin usimple equ ion estimated time o.f the deurh. The study design ol'this stuLly trus obsenational v'ilh cross-stclional approuch. The lechnique sampling used is total sanpling wilh concerning to inclusion ctnd exclttsion criterias. This shttly tvas utndttcletl Jrom ,lanuary, l't to l,lurch, 3(/h 2012 *'ith 35 people as somples. The stutly datas was obatainetl fi'om meosuremcnts o.f body temperoture in the inpatient room RSUDZA with digital tandlonrcter on the @iilla.for Lr momment. 30 minute.s and 60 minutes aJier deuth. Room temperatLtre was measto'e(l at same time. The resak ol srotistic t-test shou'ed signiJicant differences in temperature bety'een time [) minute, 30 minutes, ond 60 minutes. The restit of correlation test shoy,ed thut there is cr very slrong relationship bet*-een the decrease in body temperature as measured )t,ith digitt thermometer on the axilla to the prediction of the length o/ lime ol deoth with a correlation coelJicient oJ 0.999 which.form a linear patlern of negative andfound no platectu in lhe tarltime oJ'death. The results of this study also resulted in similarity regression of )'= 831.950 to 22.891 x thal con be ttsed to estimote the time of dedh at the beginning oJ death. Keltvords: body temperature, post-mortem, time ofdeuth