In this paper, estimates of yield and price risk of different crops in some selected regions of Bangladesh have been derived. The variance-covariance matrices of random disturbances associated with both output and prices were estimated utilising aggregate time series data in four districts. The general procedure adopted was to extract the systematic portion of aggregate time series so that the residuals represented the estimates of random components from which the relevant variance-covariance matrices were subsequently computed. Ranking of crops in terms of estimated variance of output disturbances showed that jute occupied the top position, followed by wheat, aman rice, aus rice, IRRI boro, oilseeds and pulses. This gives some idea of relative 'yield risk' of different crops in Bangladesh. Ranking of crops in terms of estimated variance of price disturbances, on the other hand, showed that pulses occupied the top position followed by oilseeds, jute, aman rice, IRRI boro and aus rice. This may give some idea of relative `price risk' of different crops in Bangladesh.