We estimate and then simulate a model of Kenyan economic development from 1965 to 1997 with two objectives in mind. The first is to demonstrate the degree of volatility of cyclical shocks that developing countries experience and to calculate the domestic nominal adjustments required by these shocks under both irrevocably fixed and free exchange rates.A comparison of these counterfactual nominal adjustments identifies the short-run implications for an economy of the choice of exchange rate regime. The second objective is to provide an estimate of the consequences for the economic development of Kenya of the lack of a coherent monetary order (excessive domestic credit expansion and overvalued exchange rate) throughout most of the period since 1965.A neoclassical convergence growth model based on Barro and Sala-i-Martin (1992) is employed and calibrated to represent the long-run growth path of real GDP in Kenya. A short-run four-sector CGE model is constructed that allows for cyclical movements of real GDP about the convergence growth path. The cyclical model focuses on the adjustment of the relative price of non-traded goods that is required to ensure short-run equilibrium in the non-traded goods sector. Given that terms of trade shocks dominated the macro environment of Kenya over the sample period, we find that a free exchange rate regime would have insulated the economy to a greater degree than an irrevocably fixed regime. In the growth decomposition exercise, we estimate that the two largest (and negative) influences on Kenyan economic growth were the decline in the external terms of trade from 100 in 1965 to an average of 79.5 over the 32-year time period, and the overvalued Kenyan shilling represented by a premium on the parallel market for foreign exchange. Overall, we estimate that the overvalued exchange rate reduced economic growth by an average of 0.47 per cent per annum over the 32 years.