10 Commonly Broken Good Laboratory Practices

What comes to mind when you think of good laboratory practices? To many, good laboratory practices describes the best conduct while working at the bench. The laboratory is a complex environment and understanding how small, seemingly innocuous, actions can have such a huge impact on the outcome of an experiment will help you to ensure that every run of an experiment is a successful run.

Think of memorable rookie research mistakes that you have made or seen others make, and you’ll have a good idea of what breaks the laws of good laboratory practice.

You might also find that you break the laws of good laboratory practice when you:

1. Wear Personal Protective Equipment (PPE) sometimes, not all the time.

Chemicals are often hazardous, and you can contaminate a sample when deciding to work with bare hands. Even if you do everything humanly possible to reduce the risk, there will always be a chance of something dangerous spilling, splashing or exploding onto your skin or clothes. Wearing proper PPE ensures that in the rare event of an accident you will be protected. PPE is inexpensive, durable and can be easily replaced, so wear your PPE every time.

2. Decide not to run a control sample.

Controls serve two very important purposes. They show whether or not your chemistry worked appropriately and they serve as the basis by which you can make a definitive comparison between groups of samples. Consider running an experiment without a control. The data that is collected is 0 overall; in essence, it appears as if the experiment did not work. Did it really fail to run, or is the data really as it is? Run a control sample every time so you know.

3. Update a laboratory notebook with only abbreviated details.

Often, you will revisit a project to review data before publication, rerun the experiment for validation or compare one experiment to another. Always making complete entries helps you to make sense of what you did in the past: “See network drive for data file” is great as long as you note which file it is… “The control sample did not work” is ambiguous until you describe the type of control sample and experimental conditions. Give yourself a helping hand and be thorough with everything you write in your laboratory notebook.

4. Don’t write anything down.

Of course, it helps to be consistent with record keeping. At the end of a project, your supervisor will want to review procedures and data from beginning to end. You can only show and tell so much at a laboratory meeting, so when it comes to anything you’ve done in the lab, one good rule to live by is, “If it wasn’t written down, it probably didn’t happen.”

5. Don’t calibrate your equipment.

An uncalibrated machine can measure fantastical values. And if you calibrate it, but select the wrong measurement mode, you can run into a situation where you grossly under- or over-measure how much stuff is inside your tube. Before using any piece of equipment, take a moment to ensure that it is properly calibrated first.

6. Use tools or equipment that are “too big” for the job.

A 100ul volume can be measured with either a 100ul or 1000ul pipettor, but the exact measurement between those two pipettors will differ. Even with a 1% error, the difference in volume pipetted could be 1ul or 10ul, respectively (+/- 10X). Every instrument has its limitations. Keep variability within your experiment low by selecting instruments that are the right size for your measurements.

7. Work through your math and units only once.

Practice the habit of double- and triple-checking your work. Before mixing up that expensive batch of media, review units and calculations to see that your numbers make sense.

8. Use samples before quality checking them.

Chemical carryover of chloroform, phenol, ethanol or salts from DNA or RNA extractions can halt reactions in other experiments. Checking the integrity and quality of your samples through spectroscopy, gels or other means is a simple way to find out if you need to clean up before moving on. Strive to generate the highest quality samples that you can, above and beyond any noted lab- and assay-minimum requirements.

9. Put off required refresher training.

Yearly refresher training may seem redundant, yet it serves a very important purpose: to ensure that all staff are on the same page when it comes to safety, conduct and responsibility. This training keeps important topics fresh in your mind. Who knows, one day you may draw upon it to help a colleague to return to good laboratory practices.

10. Communicate with your lab-mates sparingly.

The road to a successfully completed project is filled with collaboration and communication. The more you communicate with those around you, the better chance you have of accommodating everyone’s needs. This is especially important when all equipment and bench space is shared. And if you work in a lab that routinely does fluorescence microscopy, it’s always nice to have a heads-up before unexpectedly finding yourself having to work in the dark.

These 10 common examples show how good laboratory practices can be broken and why they are important to keep in mind while working at the bench. Many times recalibrating, remeasuring, or redoing an experiment is all that’s needed to return to best practices in the lab. What other conduct breaks good laboratory practice, and why is it noteworthy? Please share your experiences with us.


  1. Sau on November 17, 2016 at 2:07 pm

    That’s a great article! The laboratory practices stated above are of great importance, indeed. And if a person wants to be an expert in their job, they should really do their best not to break those practices. I wanted to improve my theoretical and practical understanding of the good laboratory practice standards, and fortunately I found an online course. After passing the exam I got a certificate, too.

  2. Talley on June 18, 2012 at 7:06 pm

    to this I would add…
    Not evaluating results blindly.

    It’s amazing how often this basic tenet of science is forgotten. Even those who claim that they can “resist bias” are susceptible to the influence of expectations. Even doing a first-pass experiment without blinding yourself (and saying “I’ll repeat it blindly later if it works this first time”) can lead to a slippery slope. No one is immune to bias, conscious or not. Blind yourself to your conditions. . .

Leave a Comment