The laboratory work involves not just knowing how to interpret the exam and help the clinician but also how to perform the pre-analytical phase, it means how to get and send a good sample to the lab, controlling problems that occur prior to sample analysis, and how to analyze the sample. There are several studies in human and veterinary medicine about this issue. Glucose and lactate are examples of analytes that can be affected by this pre-analytical phase of laboratory testing. Anticoagulants and enzymes inhibitors are used to prevent false results of glucose and lactate but information is still needed since most part of veterinary private practices in Brazil and other countries send their samples to an outside laboratory. The purpose of this study was to evaluate the effects of time of storage on canine blood glucose and lactate levels in different specimens. Fasting blood samples from 24 clinically healthy dogs were collected into four NaF/EDTA-Na₂ tubes and into four no-additive tubes (BD-Vacutainer®, São Paulo, Brazil). The samples were centrifuged in different times after collection (0, 1, 2 and 4 hours). The 1, 2 and 4h NaF/EDTA-Na₂ tubes were refrigerated (~6°C) while waiting for centrifugation and serum tubes were allowed to sit at room temperature (~25°C). Serum and plasma samples were separated and frozen (-20°C) prior to analysis. Glucose and lactate concentrations were determined by spectrophotometry. Glucose decreased with time of storage, a statistic difference was found between all serum samples (p < 0.05) and only 4h plasma NaF/EDTA-Na₂ sample had significant lower glucose level compared with the other plasma samples (p < 0.05). Between the two types of samples, serum samples centrifuged in > 1h and plasma NaF/EDTA-Na₂ kept for 4h before centrifugation had significant lower glucose concentrations (p < 0.05). Lactate increased significantly with time of storage in serum samples and there was no statistic difference between plasma samples (p < 0.05). Only serum immediately centrifuged after collection didn't differ from plasma values. No visual evidence of hemolysis was observed in plasma or serum. The authors concluded that 1h serum samples are suitable for glucose measurement in dogs, NaF/EDTA-Na₂ plasma can be used as well but 4h samples can lead to lower glucose values, probably due to glycolysis inhibition delay and dilution effects. For lactate measurement, NaF/EDTA-Na₂ was necessary to prevent the increase of lactate with time, 4h NaF/EDTA-Na₂ plasma sample and serum samples centrifuged and separated immediately after collection can be used to determine lactate concentrations.