Prostate cancer (PC) growth is driven by androgen receptor (AR) signaling, where the AR is activated by ligands testosterone (T) and dihydrotestosterone (DHT). Despite the success of initial therapy where circulating T and DHT are reduced, patients can progress to the lethal form of the disease, castration resistant prostate cancer (CRPC). Treatment of CRPC aims to stop a resurgence of AR signaling with drugs known as androgen receptor signaling inhibitors (ARSI). Unfortunately, patients develop resistance and succumb to the disease. Resistance can occur via intratumoral steroidogenesis where steroidogenic enzymes convert androgen precursors to T/DHT within the tumor. Here, we investigate a mechanism of intratumoral steroidogenesis involving aldo-keto reductase family 1 member C3 (AKR1C3), a major steroidogenic enzyme in PC, and a reservoir of the adrenal androgen DHEA-S that remains after ARSI therapy. We quantified DHEA-S metabolism using stable isotope dilution liquid chromatography tandem mass spectrometry and monitored DHEA-S induced cell growth in two PC cell lines. We found that post-ARSI concentrations of DHEA-S are converted to T in an AKR1C3-dependent manner and are sufficient to stimulate AKR1C3-dependent cell growth. We also observed metabolism of DHEA-S through the intermediate metabolite, 5-Adiol. AKR1C3’s role in CRPC could be modified by the impact of nonsynonymous single nucleotide polymorphisms (nsSNPs) that manifest as mutations in the protein. AKR1C3 variants have been associated with several diseases, but a detailed biochemical analysis to support these claims is lacking. One study reported major differences in metabolism of exemestane, an aromatase inhibitor used to treat breast cancer, by variants H5Q, E77G, K104D, and R258C compared to WT. To determine if nsSNPs could impact AKR1C3’s role in CRPC, we purified AKR1C3 H5Q, E77G, K104D, and R258C from E. coli to evaluate their kinetics parameters and stability compared to WT. We report no major differences in these variants compared to WT. Altogether, this data suggests a mechanism of resistance to ARSIs via AKR1C3’s intratumoral conversion of DHEA-S to androgens that drive PC growth and supports AKR1C3 as a therapeutic target in CRPC drug resistance, where patients could benefit from AKR1C3 inhibition no matter their nsSNP variant status.