In vitro deamination generates ammonia in freshly collected blood specimens. To prevent this, samples for ammonia testing are usually collected on ice and run rapidly (e.g., within 1h). We developed a method to stabilize specimens for ammonia analysis. Following plasma separation, 500μmol/l cycloserine or a combination of 2mmol/l sodium borate with 5mmol/l l-serine were added to sample pools with normal or increased concentrations of ALT and/or GGT to inhibit deamination; and/or residual platelets were removed via centrifugation. Sample pools were then incubated at room temperature or 4°C. Untreated sample pools were also incubated at −80°C. Ammonia was measured at 0, 1, 2, 4, 8, 16, and 24h. When incubated at 4°C without treatment, sample pools with enzymes within their reference limit had an increase of 0.5μmol/l/h, whereas sample pools with ALT and/or GGT activity above their upper reference limit had an increase of 3.6μmol/l/h (p<0.001). When sample pools were incubated at 4°C with sodium borate/l-serine, the rate of ammonia increase was significantly reduced in samples with normal (0.3μmol/l/h, p<0.001 vs. untreated controls) or high enzyme activity (0.1μmol/l/h, p<0.001 vs. untreated controls). Independent of the ALT and/or GGT concentrations, storing the sample at −80°C also preserved the specimens for ammonia analysis (0.2μmol/l/h, p<0.001 vs. untreated controls). By combining sodium borate/l-serine with refrigeration, plasma ammonia specimens can be stabilized for >12h. •We present a simple method to preserve ammonia in clinical specimens.•This method preserves plasma ammonia specimens for >12h.•In vitro ammonia increase is inhibited by 2mmol/L sodium borate+5mmol/L l-serine at 4°C.•This method works for samples with both normal and high levels of ALT/GGT activity.