Inspired by our recently developed FarmCPU method (PLoS Genetics) and its R package, we continue our practice on modeling and software engineering. One of the outcomes is BLINK written in C by Meng Huang. Both real and simulated data analyses demonstrated that BLINK improves statistical power further and deliver a remarkable computing speed. Now, a dataset with half million markers and half million individuals can be analyzed within an hour, compared with days by using FarmCPU. The executable program of BLINK can be freely used to analyze datasets up to 50 million data points (number of individuals x number of markers) in size. Releases of limitation on the data points, or Licenses for BLINK source code are available at The Office of Commercialization at Washington State University. Questions and comments can be made at BLINK forum.The executable program of BLINK, help documents, demonstration data, and tutorials are freely available as following: