This study explores the determinants of the USA's trade balance (NE) with Australia, and tests for long-run relationships. Regression results suggest the most important determinant of NE is the price ratio (USA relative to Australia), followed by the lending rate ratio, GDP ratio, money supply ratio, and the real exchange rate, which together explain 93% of the variation in NE. F-tests suggest NE responds to rising and falling exchange rate regimes symmetrically. Johansen's cointegration test suggests a significant (at 1%) long run relationship between NE and its determinants. Based on Engle–Granger, import share (IS) and output growth (Q) are cointegrated at 10% if Q shocks precede IS shocks, suggesting a weak long run relationship between them. There is a positive long run relationship between NE and the exchange rate (ER) at 5%, if ER shocks precede NE shocks. Results based on the error correction model suggest a positive relationship between expected ER and NE, and reinforce the cointegration results. The estimated NE regression model may be used to monitor and predict NE, given relevant values (predicted or predetermined) of its determinants.