Fiscal policy effectiveness in resource-dependent economies hinges on revenue and expenditure composition, yet limited empirical attention has been devoted to the disaggregated transmission of fiscal shocks in Nigeria. This study investigates the long-run and shock effects of oil revenue, non-oil revenue, capital expenditure, and recurrent expenditure on economic growth using a Vector Error Correction Model with Generalised Impulse Response Functions and Forecast Error Variance Decomposition. Key findings reveal that oil revenue exerts a negative long-run effect on GDP, confirmed by impulse responses showing persistent and deepening GDP contraction following oil shocks. Conversely, non-oil revenue and capital expenditure exhibit positive long-run associations, with capital expenditure shocks generating the largest positive GDP response and ultimately accounting for over 56 percent of forecast error variance at the twentieth horizon. Recurrent expenditure shows no reliable growth relationship; its variance contribution declines sharply over time. Implications underscore the imperative of sterilising oil windfalls, prioritising non-oil revenue mobilisation, and insulating capital budgets from revenue volatility to enhance long-term growth outcomes in Nigeria.