ÖZE TFiyat, verim ve üretim gibi zaman serileri sistematik(trend) ve tesadüfi (random) olmak üzere iki bileşene ayrılabilmektedir.Zaman ser/sinden belirli yöntemlerle ayrılantesadüfi bileşeııden hesaplanan bazı lstatistikler (standart,sapma, değişkenlik katsayısı gibi) söz konusu zaman serisineait belirsizlik ölçüsü olmaktadır. Araştırmada 1950 - 1973dönemiTürkiye buğday verim serisi ele almarak trend. trenddensapmalar ve otokorelasyon testleri yardımiyle tesüdüjibileşen ayrılmakta ve bu bileşenden hesaplanan değişkenlikkatsayısı buğday verimine ait belirsizlik ölçüsü sayılmaktadır.SUMMARYEstimation of Random Fluctuations and of the Measure of Uncertainty in Time Series Analysis.Time series such as price, yieldand production general1y can be separatedinto two components : the predictableco;ınponent, which includestechnological advances and long-runeconomie trend, such as inflation andpriee cycles ; and the unpredictableconıponent, which is nonnally associatedwith weather. When planning forthe current years crop program, far;ıners are probably more interested intheunpredictable variability affectingcurrent levels of prices, yields, and inco;me. it is only this unpredictable, orrandom co;ınponent which is tried tobe;measured in this study.Wheat yield per hectare in Turkeyfor the period of 1950-1973 is takenas a samp1e in this study. By eertaintrend and autocorrelation tests, therandom co;ınponent is separatedfrom the series. By using this randomcomponent and the 1950-73 averagewehat yield, a coeffident of rando;mvariation which is regarded as a mesUreof uncertainty, is computed. For exa;ınple, the random variation (variationother than long-term trend) in wheatyield over the years has been 154.fS kg.per hectare. Averege wheat yield forthe period of 1950-1973 is 1124.71kg. per hectare. Accordingly, the co·efficient of random variation for wheatyield is 154.15/1124.71 = 0.1371, or13.71 per cent.