Bone resorbing osteoclasts are specialized macrophages that cannot differentiate in the absence of c-Fos, a member of the dimeric transcription factor AP-1 (activator protein-1). However, osteoclast differentiation in the absence of c-Fos can be rescued in vitro and in vivo by Fra1, a Fos-like protein and transcriptional target of c-Fos. To enable AP-1 proteins binding to DNA, c-Fos or Fra1 must heterodimerize with a partner such as c-Jun, JunB and JunD. In this study, we investigated the dimerization partners of c-Fos and Fra1 required for osteoclast differentiation using synthetic "single-chain" AP-1 dimers in which c-Fos or Fra1 is tethered via a linker to Jun proteins. When c-Fos was analyzed in combination with any Jun protein, including a c-Jun mutant lacking major phosphorylation sites for c-Jun amino-terminal kinase (JNK), osteoclasts were efficiently formed from c-Fos-deficient hematopoietic precursors. However, Fra1 in combination with any Jun protein could not rescue osteoclastogenesis. The ability to rescue was compared to transcriptional activity measured in transient transfection assays using promoters driven by consensus AP-1 sites or a composite AP-1/NFAT binding site. These data show that a single Jun/c-Fos dimer is sufficient for osteoclast differentiation, likely due to its transactivation ability for a broader range of promoters, in particular consensus AP-1 sites. We propose that Fra1 together with a dimerization partner different from Jun proteins can rescue osteoclast differentiation in c-Fos-deficient precursors.