Sample collection, transport and storage conditions vary in the human cytomegalovirus (CMV) shedding literature. Currently, limited data exist on the impact of biological fluids and pre‐analytical sample handling on the detection of CMV DNA. To evaluate CMV DNA recovery from urine, vaginal fluid and saliva stored in different conditions, adult urine, vaginal and saliva fluids and swabs, stored with or without selected nucleic acid preservation media at various durations and temperatures, was compared by polymerase chain reaction (PCR) quantitation of spiked samples and self‐collected urine (n = 45) and vaginal swabs (n = 58) from CMV seropositive pregnant women. There was a time‐dependent reduction in CMV DNA recovery from urine, urine diluted in phosphate‐buffered saline, and saliva stored at 2−8°C, but not from urine preserved in cobas® PCR transport media (CPM) (urine/CPM). For vaginal fluid, a reduction in recovery was evident after 7 days storage at 2−8°C. CMV DNA recovery over 91 days was similar between −80°C and −20°C storage for urine and vaginal swabs preserved in CPM, and saliva swabs preserved in eNAT® PCR transport media. A statistically significant change in CMV DNA recovery after 25 months storage (median) at −80°C was not observed for self‐collected urine/CPM and vaginal swab/CPM from pregnant women. Taken together, recovery of CMV DNA is dependent on fluid type and storage conditions. To improve the validity and reliability of detection at different storage durations and temperatures, the use of nucleic acid preserving transport media at the point of collection for urine, vaginal fluid and saliva may be essential.