Immunoprecipitation (IP) coupled with LC-MS/MS is a widely used method in proteomics research to identify proteins and to study protein-protein interactions. IP techniques allow purification of proteins of interest and reduce sample complexity before introduction to the mass spectrometer. The effectiveness of IP experiments is an important factor for identification of proteins and protein-protein interactions. In this paper, a variety of IP conditions were studied systematically to improve IP-based protein interaction identification capabilities. Low concentration detergent (around 0.05% NP40/PBS) was found to improve IP effectiveness by decreasing non-specific binding. However, higher concentration detergent (e.g. 1% NP40/PBS) was detrimental. Furthermore, with lower protein concentrations, the IP system showed lower tolerance to detergent. For example, with a detergent concentration higher than 0.05% NP40/PBS, IP experiments were unsuccessful with low protein concentration (e.g. 0.28 µM ADH). In some cases the observed results were even worse than the results obtained without detergent. However, when the protein concentration was high (e.g. 1.12 µM ADH), this effect was not obvious and the high detergent (higher than 0.1%) experimental results were similar to those from low detergent concentration experiments (around 0.05%). Another application of this strategy to a more general system based on FLAG-Bacterial Alkaline Phosphatase (BAP) and anti-FLAG antibody was also performed. These results also suggested that low detergent concentration (0.05% NP40) is helpful for IP experiments, especially for the experiments with low protein concentrations. Considering that in most real applications, the proteins of interest are usually present in low abundance, a low amount of detergent is recommended to be used. The optimized detergent concentration was determined to be 0.05% in these studies. However, the key result presented here illustrates that both detergent and protein concentrations should be carefully considered when one is trying to optimize IP prior to mass spectrometry experiments.