Microfluidic devices have been used in various biological experiments. The working temperature of the devices spans a wide range (approximately 23 °C–95 °C). Among thermoplastic materials, cyclo olefin polymers (COPs) are promising materials for microfluidic devices. This is because COP can overcome the well-known disadvantages of polydimethylsiloxane, a commonly used material, and have the advantage of better observability than polystyrene and polymethyl methacrylate. However, most COP-based devices are fabricated using solvents and adhesives during the bonding process. These solvents, which are known to affect biological experiments, may remain in the device and be released during the experiments. It is necessary to investigate whether solvents are actually released and, if so, how they are released. Here we introduce thermal desorption spectroscopy as a simple and quantitative method to observe solvent release from solvent-bonded and vacuum ultraviolet (VUV)-bonded products. Solvents are released from the solvent-bonded product at 31.5 °C, suggesting that it may have negative effects on various biological experiments. On the other hand, the VUV-bonded product releases solvents (cyclohexane and toluene), which are used during olefin polymerization in the synthesis process of COP, at temperatures above 84 °C. Therefore, the experiments conduct below 84 °C (e.g. in situ hybridization, reverse transcription (RT) and loop-mediated isothermal amplification) were not affected. In addition, the amount of solvent released above 84 °C is small (1/548–1/913 of the solvent-bound product), so it is expected that the extent of the effect on experiments conducted above 84 °C (RT and polymerase chain reaction) is small, if there is any. We conclude that solvent-bound devices can have undesirable effects in many biological applications, not just cell culture. We believe that evaluating solvent release from devices is important for the development of new devices in the future.