Matching format and sample rate matters more than people expect
When source clips have different sample rates, bit depths, or formats, joining them requires normalizing them to a common set of properties first — skipping this step can produce pitch shifts, speed mismatches, or outright playback errors in some players, since a file's audio properties aren't just metadata but define how the raw samples are meant to be interpreted. This is the single most common reason a join goes wrong.
The safest approach when you're not sure whether your clips match is to let a tool normalize everything automatically before joining, rather than assuming files from different sources (different recording apps, different downloads, different original formats) are already compatible.
The join point is where all the artifacts live
Just like cutting, joining audio at a point where the waveform isn't near zero amplitude can create an audible click at the seam — and because you're now dealing with two waveforms meeting rather than one being cut, there are two opportunities for this to happen: the end of the first clip and the start of the second. A clean join needs both edges to land in a natural gap or near a zero-crossing.
A short crossfade — even 10-50 milliseconds — at the join point smooths over small mismatches at the seam that a hard cut-and-paste join would expose, and is standard practice in professional audio editing specifically for this reason.
Order matters, and it's easy to get wrong with many files
When joining more than two or three clips, it's easy to lose track of the intended sequence, especially if filenames don't sort in the order you want them played. Double-checking the final order before committing to the join — rather than after — saves having to redo the entire process, since reordering after joining usually means starting over rather than a quick fix.
Joining doesn't have to mean re-encoding everything
If all source clips already share the same format and encoding settings, joining them can often be done by concatenating the underlying data directly, without a full re-encode — preserving whatever quality each clip already had rather than introducing another compression pass. This is only possible when the clips genuinely match; mismatched clips require re-encoding to a common format as an unavoidable step.