This paper considers the situation in the laboratory testing: different stress types and accelerated testing, including accelerated reliability/durability testing, accelerated life testing, reliability testing, proving grounds, vibration, temperature, voltage, humidity, and others. In comparison with field situation, most of these testing simulate only one or part of the field input influences. One uses often not accurately the theory of physics-of-degradation process or failures for comparison of the field results with laboratory results. Thos situation will be considered with practical examples. It will be demonstrated that often used laboratory testing does not offer the possibility for successful prediction of product performance during service life As a result, there are many complaints, recalls, less profit than was predicted during design and manufacturing. It will be shown how one can improve this situation..