Background Studies regarding genetic and clinical characteristics, gender preference, and gonadal malignancy rates for steroid 5-alpha-reductase type 2 deficiency (5α-RD2) are limited and they were conducted on small number of patients. Objective To present genotype–phenotype correlation, gonadal malignancy risk, gender preference, and diagnostic sensitivity of serum testosterone/dihydrotestosterone (T/DHT) ratio in patients with 5α-RD2. Materials and methods Patients with variations in the SRD5A2 gene were included in the study. Demographic characteristics, phenotype, gender assignment, hormonal tests, molecular genetic data, and presence of gonadal malignancy were evaluated. Results A total of 85 patients were included in the study. Abnormality of the external genitalia was the most dominant phenotype (92.9%). Gender assignment was male in 58.8% and female in 29.4% of the patients, while it was uncertain for 11.8%. Fourteen patients underwent bilateral gonadectomy, and no gonadal malignancy was detected. The most frequent pathogenic variants were p.Ala65Pro (30.6%), p.Leu55Gln (16.5%), and p.Gly196Ser (15.3%). The p.Ala65Pro and p.Leu55Gln showed more undervirilization than the p.Gly196Ser. The diagnostic sensitivity of stimulated T/DHT ratio was higher than baseline serum T/DHT ratio, even in pubertal patients. The cut-off values yielding the best sensitivity for stimulated T/DHT ratio were ≥ 8.5 for minipuberty, ≥ 10 for prepuberty, and ≥ 17 for puberty. Conclusion There is no significant genotype–phenotype correlation in 5α-RD2. Gonadal malignancy risk seems to be low. If genetic analysis is not available at the time of diagnosis, stimulated T/DHT ratio can be useful, especially if different cut-off values are utilized in accordance with the pubertal status.