The four artifacts of a marriage
A typical US marriage generated up to four records: the bond (pre-1850s, especially in the South), the license or application, the return or certificate, and the church register entry. The bond was a financial surety guaranteeing that no legal impediment existed; the applicant and a bondsman signed it. The license authorized the ceremony; the return confirmed it had been performed.
The most useful data lives in the application: full names of bride and groom, their ages, residences, parents' names, the bride's maiden name, and occupations. In the later 19th and 20th centuries, applications also carry places of birth and previous marriages. The church register adds the officiant and witnesses.
Reading the abbreviations
Ages are often written as of aet. or simply followed by y. for years. A second marriage is sometimes flagged 2d. m. or by 'wid.' (widow) / 'wdr.' (widower) next to the spouse's name. Ux. appearing beside a woman's name in a later census or probate means 'wife of', tying her back to a specific husband when no maiden name is recorded.
Common catches
Marriage records are jurisdiction-specific. Bride and groom often traveled to the bride's home county, which is not necessarily where either lived after the marriage — check adjacent counties and any known urban center nearby. If you find a bond or license but no return, the marriage may not have taken place; the couple may have eloped elsewhere, or one party died between license and ceremony.