Temperature variables change continually through time, and the data exist as time series. It is consist of random error components with stochastic variations in space. Parametric Vector Integrated Autoregressive and Moving Average models through Monte Carlo simulation method were successfully applied to model trend of average monthly maximum temperatures and the average monthly minimum temperatures time series on Cairo governorate; Egypt, during January 1961 to December 2007. Spectral analysis was used to correct the data for cycles and periodicities. First differencing of the series provided the required seasonality, and non-stationarity corrections. Model orders and model structures were identified through spikes and lags in the autocorrelation, partial autocorrelation, and inverse autocorrelation function plots of the data. Fitted models allowed multiple, input and, output series and produced excellent fit of the data. Model parameters were significant with non-significant correlations at 5% level of significance. Unexplained variations were within 10% of average monthly maximum temperatures and average monthly minimum temperatures. Goodness-of-fit statistics indicated adequate statistical fit of the model. Root mean square errors of the models were less than 0.3. No autocorrelations were detected in the residuals estimated from the models. Time series modeling of average monthly maximum temperatures and average monthly minimum temperatures trend using integrated stochastic models is simple, cost effective and powerful.