Thoughts on epoxy vs splicing new wood for rotted door jambs?

I tried the filler shortcut on my side garage door jamb about four years ago and regretted it. Regular Bondo doesn't expand and contract at the same rate as the surrounding pine, so after two Midwest freeze-thaw cycles, a hairline crack opened right at the joint, let moisture in, and popped the chunk loose.

For anything three inches deep, splicing in a Dutchman is relly the better route. Pull those old nails out with locking pliers or buzz them off flush with an oscillating tool. Make a clean, slightly upward-angled cut on the existing jamb so any moisture naturally sheds outward, then rough-cut a replacement block out of cedar or treated wood. I butter the mating surfaces with Titebond III or a smear of thickened epoxy, countersink a couple of exterior screws to lock it into the framing, and then just use a tiny skim of exterior filler over the seam to smooth it out before paint. The most important detail is leaving a 1/4-inch gap between the bottom of your new wood and the sill or concrete. Coat that bottom end grain heavily with primer or epoxy before you screw it in, otherwise it will just wick up standing water and rot out again in five years.

I have to disagree on splicing in fresh wood. If the bottom three inches turned to mush, moisture is clearly pooling right at that sill plate, and putting a pine Dutchman right back in the splash zone is just setting a five-year timer on doing this job all over again.

What worked much better for my back patio door was cutting the jamb off clean about five inches up using an oscillating saw, pulling those rusted nails, and splicing in a block of cellular PVC like Azek. It cuts, routs, and takes paint just like real wood, but it will physically never wick water or rot out. I bedded the top seam in polyurethane construction adhesive, pre-drilled two stainless trim screws diagonally into the framing, and kept the very bottom 1/8 inch off the sill before hitting it with OSI Quad caulk.

If you really don't want to cut higher and prefer filling, don't write off epoxy just because auto Bondo failed for someone else. Bondo is rigid polyester that repels nothing and snaps off wood. Real architectural epoxies like Abatron WoodEpox actualy expand and contract with seasonal changes, but composite at the base is still the most bulletproof route.

Whatever you replace it with, don't forget about the weatherstripping slot. I patched my front entry jamb three years ago and completley forgot to cut that kerf into the replacement piece before screwing it to the stud. Trying to cut that tiny slot with an oscillating tool while crouched on the concrete porch was pure misery. Now if I have to make a patch, I measure the groove on the old jamb and cut it on the table saw before installing. Saves a ton of headache.

Before deciding on filler versus splicing, have you acutally ruled out rot behind that jamb?

Three inches of mush ussually means moisture has been sitting there for a good while. Have you poked around the jack stud or the subfloor underneath with an awl yet? If the framing back there is spongy, patching the jamb profile is kind of just putting a band-aid on a bigger problem. Might be worth probing around back there first before you spend time fitting any wood or filler.