The primary purpose of the dissertation is to establish a conceptual study that will assist in financial policy formulation, and further applied research on banking institutions in LDCs and in Nigeria in particular. In order to achieve this, we examined existing models of banks' portfolio behaviour with a view to isolating major issues that should be the concern of policy; the pecularities of banking operations in LDCs were also examined and analysed. Having done these we then settled for two adaptable models capable of providing insights into the problem of commercial banks' portfolio behaviour in Nigeria. The choice of models was made in the light of the data constraint faced in the study. The two complementary models, the first a static model and the other a dynamic model have their basis in utility maximization. They both gave rise to multivariate analysis in empirical implementation. The summary of results with policy implications concern: (i) the relative insignificance of the interest rate variables and their relative inelasticities, particularly with respect to bank loans and other medium term bank portfolios; (ii) availability of funds were more significant on banks portfolio allocation of funds than other variables, particularly those measuring the cost of funds; (iii) on other policy variables, reserve policies of the financial authority were important in bank portfolio, whereas public sector borrowing from the banks were less important; (iv) the term transformation implied by the portfolio of the banks was shown to be less positive than the underlying structure of funds would imply; (v) the introduction of money market instruments were shown to have affected the adjustment behaviour of the banks; and (vi) the results also held implication for the control of money supply. However, we believe the major contribution of this study to knowledge lies in indicating the responses and lags of banks' behaviour to changes in policy instruments in a developing country like Nigeria. This information would no doubt form the basis for further research.