Objective: To study the pregnancy outcomes of conservative treatment for preeclampsia with severe features. Methods: A retrospective study was conducted on pregnancies with preeclampsia with severe features at gestational age 23–34 weeks and that received conservative management at Chiang Mai University Hospital between January 2014 and August 2020. The women were divided into two groups: (1) pregnancy prolongation of at least 48 h and (2) pregnancy prolongation of less than 48 h. Results: Of the 100 recruited pregnancies, the median gestational age was 29 weeks (range 23–34). Of these, 65 cases (65%) had pregnancy prolongation of at least 48 h, and 35 cases (35%) had prolongation of less than 48 h. The median pregnancy prolongation was 2.9 days (range 4 h–27.7 days). Eighty-seven (88%) pregnant women experienced no complications. Multivariate analysis shows that high urine protein/creatinine ratio (UPCI) at admission was significantly associated with pregnancy prolongation of less than 48 h with an odds ratio for prolongation for at least 48 h of 0.86 (95% CI 0.75–0.99: p-value 0.04). Kaplan–Meier analysis shows that the mean time of prolongation was 3.6 days vs. 6.7 days, and median time of prolongation was 2.1 days vs. 4.4 days in the group of high and low UPCI (using cut-off 1.0), respectively. The number of prolonged days was significantly lower in the high UPCI group than in the low UPCI group (log-rank test, p = 0.01). The maternal and fetal outcomes between the two groups were not significantly different. The cesarean section rate was also comparable. The mean birth weight and gestational age at delivery were not significantly different, though they had a higher trend in the group of successful conservative management. Conclusion: The rate of pregnancy prolongation of at least 48 h with conservative management was 65%, with a median prolongation time of 2.9 days. A new insight gained from this study is that high UPCI at admission is an independent factor for prolongation of less than 48 h with conservative treatment. Nevertheless, the maternal and fetal outcomes between the two groups were not significantly different. Therefore, the benefit and risk of expectant management in actual practice of service settings in terms of maternal and fetal morbidity is still unclear.