In vitro experiments in the early 70s suggested that the time it takes for LacI to bind to the operator site is ∼100 times faster than what would be expected for a pure 3D diffusion-controlled binding reaction. This discrepancy between theory and observations was resolved by introducing the theory of facilitated diffusion where the operator search problem is viewed as a coupled 1D-3D diffusion process. Since then, numerous theoretical studies have been made to elucidate the details of the search process. In this work we have used fully atomistic molecular dynamics simulations for the complete LacI dimer (666 aa) and a stretch of nonspecific DNA (30 bp) to help unravel these details. In particular we find that LacI slides along DNA in a helical trajectory with fast hydrogen-bond dynamics, displacing water molecules and sodium atoms in its path. Using the umbrella sampling technique we also quantify the free energy variation of LacI sliding and dissociation. We then use this energy landscape to deduce the 1D diffusion constant, D1=0.14 μm2s−1, and a microscopic residence time, τmicro= 69 μs. Given that τmicro is not accessible on experimental timescales we employ Brownian dynamic simulations to estimate the average number of times LacI rebinds to DNA before macroscopically dissociating (∼700). Then, by connecting Brownian dynamics and molecular dynamics, we get a macroscopic residence time (τmacro= 48 ms). Finally, we estimate a sliding length of 240+-105 bp, which compares well to the experimental value of 150 bp. As such, we connect structural dynamics at the ns timescales and atomistic lengthscales with meso- and macroscopic information to unravel hitherto unknown apects of the search process.