The inactivity or occlusion of the HIV-1 poly(A) signal when in the 5′ long terminal repeat (LTR) has been mechanistically investigated. First we show that neither the homologous HIV-1 promoter nor the close proximity of this RNA processing signal to the transcript initiation site is required for the occlusion effect. Instead we demonstrate that the major splice donor (MSD) site positioned about 200 bp downstream maintains the poly(A) site in an inactive state. Although mutation of MSD results in activation of the 5′ LTR poly(A) signal, this effect can be suppressed by targeting U1 snRNAs near to the mutated MSD by base pairing. We show that hybrid U7-U1 snRNAs can also suppress the poly(A) signal and that this suppression is dependent on the U1 stem-loop 1. In particular the binding site for the U1 snRNP protein 70K that binds to the loop structure of stem-loop 1 is associated with poly(A) site occlusion. These experiments were carried out with an HIV-1 proviral construct and as such emphasize the physiological importance of this splice donor–poly(A) site interaction.