Short answer is maybe. However, if it may not be accurate, then should it really be used against you at trial? The State of Texas says yes. Admittedly, the Intoxilyzer 5000 (5000) is quite good at measuring the reference sample (closed canister containing water mixed with a known quantity of alcohol that is connected to the 5000).
A few problems arise when the principle used to determine alcohol concentration in the reference sample are used when trying to measure a breath sample from a human being. Henry's Law says that in a closed container, with a constant pressure, constant temperature, and a fluid with airspace above, the fluid and air will arrive at an equilibrium, and one of the major assumptions used in breath testing is that there is a correlation of 1 to 2100 between the amount of alcohol in a person’s exhaled breath and the level of alcohol in the person’s blood. (Breath alcohol is used as a surrogate measure for blood alcohol concentration.) This so-called “partition ratio” is based on averages derived from scientific studies. As noted, the reference sample is a known ratio, but in the human body the ratio can vary from 1 to 900 ratio to a 1 to 3500 ratio. This means that if you are tested and your ratio is closer to 1-900, then the 5000 will overestimate the actual concentration in your blood by up to 50% and may make an innocent person look guilty. Because of this problem, Texas law was changed to say that any score over .08 means you are intoxicated, even if a blood sample taken at the same time would have shown you were under .08. Texas law does allow you to obtain your own blood test within two hours of taking the breath test, but you have to arrange for a nurse to come and take the sample. How you are expected to find a nurse in the middle of the night is unknown.