Cross-modal　remote　sensing　image-audio　(RSIA)　retrieval　aims　to　use　audio　or　remote　sensing　images　(RSIs)　as　queries　to　retrieve　relevant　RSIs　or　corresponding　audios.　Although　many　approaches　leverage　labeled　samples　to　achieve　good　performance,　the　performance　cost　of　labeled　samples　is　high,　because　cross-modal　remote　sensing　(RS)　labeled　samples　usually　require　huge　labor　resources.　Therefore,　unsupervised　cross-modal　learning　is　very　important　in　real-world　applications.　In　this　article,　we　propose　a　novel　unsupervised　cross-modal　RSIA　retrieval　approach,　named　self-supervision　interactive　alignment　(SSIA),　which　can　take　advantage　of　large　amounts　of　unlabeled　samples　to　learn　the　salient　information,　cross-modal　alignment,　and　the　similarity　between　RSIs　and　audios.　Since　self-supervised　learning　lacks　the　supervision　of　label　information,　we　leverage　the　similarity　between　the　input　RSI　information　and　audio　information　as　the　supervision　information.　Besides,　to　perform　cross-modal　alignment,　a　novel　interactive　alignment　(IA)　module　is　designed　to　explore　fine　correspondence　relation　for　RSIs　and　audios.　Moreover,　we　design　an　audio-guided　image　de-redundant　module　to　reduce　the　redundant　information　of　visual　information,　which　can　capture　salient　information　of　RSIs.　Extensive　experiments　on　four　widely　used　RSIA　datasets　testify　that　the　SSIA　performance　gains　better　RSIA　retrieval　performance　than　other　compared　approaches.