I do, and I have three reasons for it:
1. Efficiency. C generates little surprising and/or additional code, even if you aren't exactly extremely familiar with your compiler. This allow you to think in terms of assembly instructions and clock cycles for an embedded software project, but you don't have to really mind the minute details like register allocation, ABI and even actual assembly mnemonics. And in the corner case when it arises, you can inline assembler directly. Also C has a tiny runtime, which means code is small and efficient.
2. Portability. For higher level programs I still prefer C over other languages, as it provides excellent portability between various operating systems. My daily driver machine runs macOS on Intel, and some of my projects target Raspberry Pi. As long as I stick to C + POSIX-compliant libc + a few common dependencies like libcurl, libfastcgi and nginx, I can expect my C code, debugged on my macOS workstation, work as designed on the Pi after just a recompile. I can even switch out the compiler if I want (clang on macOS, gcc-9 on Pi.) Don't get me started on C++ as that thing has such a fragile ABI. (There is a reason why Apple now implements the macOS C++ ABI on top of the Objective-C ABI, as that is just more reliable.)
3. Control. C gives me full control of the program flow - no garbage collection throwing timings out of the window, no unexpected initializer calls, no confusing operator overloads, etc.