It depends on the treatments, conditions... you are using/ testing in your experiment. You have to check if your reference gene is not affected/ stablely expressed under  those different conditions. A reference gene that is often used for example is actin2

For the plant field you could check this publication dealing with different reference genes: https://www.researchgate.net/publication/7597407_Genome-Wide_Identification_and_Testing_of_Superior_Reference_Genes_for_Transcript_Normalization

Article Genome-Wide Identification and Testing of Superior Reference...

For mouse and human samples, I have been using GAPDH gene for normalization.

It depends on your cell type, too. For human samples, i always use the RPL13a gene. I think it can be interesting for you to use a reference gene which is expressed in all the cells (as RPL13a) and a gene which is specific to your cell type (for example, for my liver cells, I use LMF2). Good luck !

For quantitative PCR, you should be using at least 3 references.  I prefer one ribosomal gene, one structural gene (like actin), and a third gene that should not change in your experiment (GAPDH is popular, but changes in many experiments, so it must be compared to the other references). As Camille said, the reference gene depends on your cell type; it also depends on your experiment, tissue, model, and treatment conditions.  References can fluctuate with circadian rhythms, in response to drugs, between different tissues, and definitely between models.

GAPDH is good. I strongly advise against any cytoskeletal genes, as I've seen big changes in those supposedly stable 'housekeeping' genes again and again. Multiple housekeepers is indeed a good idea. HPRT is pretty good as well.

Hi, we are successfully using RPL32 (L32) as reference gene for more than 10 year.

If you are using the yeast Saccharomyces cerevisiae, here is a good reference for you.

In my lab we commonly used GAPDH as a reference/control gene, we used to multiplex our genes using the GAPDH primer-probes marked with another fluorochrome different from the one for the target gene/s (FAM for the target gene/s primer-probes and VIC for the GAPDH, for instance). Also, the rRNA 18S was commonly used, although from my experience it worked better with human provenance samples (cells or tissue) than with mice ones.