This Article reports the result of a computational study on the reaction of hydrazoic acid and trimethylindium (TMIn), coadsorbed on TiO 2 rutile (110) surface. The adsorption geometries and energies of possible adsorbates including HN 3 -In(CH 3 ) 3 (a) and its derivatives, HN 3 -In(CH 3 ) 2 (a), N 3 -In(CH 3 ) 2 (a), N 3 -In(CH 3 )(a), and N-In(a), have been predicted by first-principles calculations based on the density functional theory (DFT) and the pseudopotential method. The mechanisms of these surface reactions have also been explicitly elucidated with the computed potential energy surfaces. Starting from the interaction of three stable HN 3 adsorbates, HN 3 -O b (a), H(N 2 )N-O b (a), and Ti-NN(H)N-O b (a), where O b is the bridged O site on the surface, with two stable intermediates from the adsorption and dissociative adsorption of TMIn, (H 3 C) 3 In-O b (a) and (H 3 C) 2 In-Ob(a) + H 3 C-O b (a), InN products can be formed exothermically via four reaction paths following the initial barrierless In-atom association with the N atom directly bonded to H, by CH 4 elimination (with ∼40 kcal/mol barriers), the InN-N bond breaking and the final CH 3 elimination or migration (with <20 kcal/mol barriers). These Langmuir - Hinshelwood processes producing the two most stable InN(a) side-on adsorptions confirm that HN 3 and TMIn are indeed very efficient precursors for the deposition of InN films on TiO 2 nanoparticles. The result of similar calculations for the reactions occurring by the Rideal - Eley mechanism involving HN 3 (a) + TMIn(g) and HN 3 (g) + TMIn(a) indicates that they are energetically less favored and produce the less stable InN(a) with end-on configurations.